Skip to main content
Loading Explore guides…
Nastran Toolkit · Acedra
Home
Explore
Engineering
Nastran Toolkit
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
652 segments
Copy
Hide video
0:02
welcome good morning or everybody there's time zones good whatever it is I'm filling out applications nearby yeah he transfer Montreal been with my for 18 years and before that was in 410 I think you want interested Maya yes there was a little modest he says he's an application engineer my fila disap alam i been around for a long time he's an application engineer he's a software developer he's a product
0:45
manager he's a shocker primarily responsible for sa TK they do a great job for us that are on the line that don't know anything about Maya we have real siemens PLM partner are how you add a reseller even we're also integrated socrata vel so many of the tools that are available in in the Siemens and a kind of CAE products a lot of them would probably got about sixty modules that we that we developed for Siemens the city front so join the hip with Siemens but what we're
1:33
talking about to you guys about today is the strong alice's toolkit for nastran and as went through an extensive evaluation on NX response versus SAT K and I think everybody's probably seen the numbers internally any proud of this project we hope get you to revise the benefits of it so here today for to do probably almost a four-hour trench session just an introduction to SAT K earlier there's some elements of SAT K that you guys really don't even know about you've got
2:21
full capabilities you've got the basic module you got the advanced modules you Princeton focusing on the random aspect of it but what the fields going to try to cover today at least reduce you to some of the other tools of SAT K that may benefit you that you don't even know that you got right now so anyway so let's start this off we got probably the next three three and a half 345 to go and like I said five training today this is pretty much a download is a whole lot
3:04
of interaction there's not tutorials but we we make available the temporary licenses to put on your seal server that guys can access if you want to run your Llorona Europe here in the near future so bill and octopus is with us he's a he's a developer of the class matter of este toolkit so it's really his baby and I think he's done an outstanding job and I guess the performance numbers you've shown is substantiate that um and so my fish I have about hundred twenty
3:44
slides six demos you don't need a license there are no tutorial this morning as Mike said will be some additional licenses available on request able to to try out the code yourself on your own problems there are example problems we'll discuss that later so many questions as we go along I'm a fell down because I tend to go too fast so this is what we're going to be covering thank anybody's as smart as you oh they're smarter than that how so the contents of today's presentation
4:26
is we're gonna look at this a toolkit main user interface and then we're going to look at four of the tools there's a tools and all that come with the other product what what we call dynamic tool so we look at a random vibration processor does a performance a a base excitation a frequency domain for random vibration and then from a solution 103 nasty and run and then look at the sign process does a similar type of job but with our harmonic excitation animals that we use
5:02
here in our own work to do two sorts of work and we use the tool quite often so we will understand understand several content is of our structures using the modal processor in the energy process this force that's come with the toolkit for the processors and if you're interested we can go over them at the end or when I'm done with these the main presentation we said you
5:36
interesting I'll just go to brief overview and I'll give you what's available for you so the house is toolkit how this came about to be was a previous incarnation it's farmer space which is now MBA we we feel that nasstrac was not and Patra and whatever we're using at the time wasn't quite cutting it in terms of our of our needs so yeah learned in your own code and then we joined money we redid the everything at better and code extensively here so originally
6:15
it was meant for to be to get into fee map and now it's really it's a standalone tool so mike minute we are a Siemens development part we a lot of our picks are you know have seen his name on them not this one says this is very this Maya this is a Maya products that alone I'm irate it's getting some closer to EDX now so yeah showing you how to create the input files in EDX and post-processing the results some of the results in 10x but it does
6:49
not create an X license such as mesh for it is a Maya license server which you must download
6:57
select the product so if you look at the main dialogue there's there's a button open o P to file so you're not icing solutions were we're accessing directly o P 2 files so when would it not be - for a while I'm clear if you've done a lot of modal analysis pretty much the same as you would have already done and now you do the post-processing stresses random elaboration they don't be too file he's opening it up now yeah there's other
7:40
processors of course that don't use the modal a results but they basically all the processors use master and results to stop with these a static or a transient or this case more
7:57
would go and open the op2 file that you want to process we support MSC nesrin & NX master and will be two files you guys using NX some you know this there's a variety of formats for op2 files we before see the the default format amount of NX if you start with those you say I'd be able to read the cop12 toolkit but all you what the requirements are for toolkit to process the op2 files in terms of a string and a syntax by many are you are familiar with what's your familiarity
8:36
with that next master and is it a beginner is it like if you and the syntax of the input file
8:53
I'm moderate on my end of everybody as a beginner online okay well I'll do some some of the car we call them cars or some of the keywords that of the input file that that tool gets exciting so you don't necessarily need to go into everything works out fine and then you don't really about that but if you know you hit a snag then it's good to understand what is written to the some of these keywords are and and what the reports are but it's okay don't be too far make dialogue to
9:36
a part actually some of you guys should I hear some background noise maybe it's best to go to mute yourself Oh what the what the toolkit does is dialog is it Parsons EOP to file it shows you the data blocks the master and data blocks that are in it and you can see some of them listed listed here um so that's been and then what we found is that for some really large will be too far I'll
10:04
I'm you know a minute maybe two minute so we we did quick open I wasn't scan you'll be to file so we don't know the blocks are available it was you all the dynamic processors and there's something missing in the op2 file then the error there be trapped by the processors themselves sighs I'm no data blocks the provider will be available to you anyone what's in them so that if you burn and
10:42
then you go to tools you'll see which processors okay use if use quick open then you basically have to know what you're doing there's a prod can set so by default the project defined the the foot file names and I default it's the UH it's the op2 file name with a subscript which coupons to the processor that you're using but you control over that you can name the project whatever you want
11:13
the next the nectar menu item groups are porn because they allow you to generate margins of safety and in particular or with with the processor or post process just certain part of your model you can rate the groups yourself who knows the relevant UI logs eels that you see when your different output requests or you can forth as I before we essentially started building this for fee Maps you see a lot of references for fee map but in our case which we want to do here from
11:51
NX is import those groups from a universal file and I'll show you how to create the groups how does sorry export the groups in Universal file format so yes but you can if you warm
12:11
this is a bit confident than the future we you know certainly going to on that but I'm your when you nnx you have these lenses or these solvers that you pick when you beta FM which is to change your self so in this indicator I like your FEM you edit the fem and then you your solver from an X and a strand to ideas you envy
12:40
and then you sports simulation from the NX main menu bar and what that do is connect for the Intolerable into the universal file including the groups that proves that you define and have yet here well first the groups must contain and/or element so this is really what toolkits using it's not using in your entity NX by default try to put only muscies in the group's because it's simpler than putting all the elements in the mesh it's more
13:14
it's lighter but you got to make sure that you have the elements as well not the mesh but the elements so I'll show you how to do that if you're not familiar with thanks group function you know
13:30
the one supported once you selected you will be to file and once you've imported your group um you can see that they're available as I mentioned before if you open a we be two files so only the tools that are set sick reference to be the information you'll be to file are available yeah I click open the no dynamic tools are available though that energy random modal and sign
14:01
you
14:04
Oh menu here is launches a dialogue which is Blee informational it shows you all the files you've created for the project you can't edit them you can do them so this is something that you know we've I guess except of energy we have pushed really to as far as we could but this is something that is not as useful as it could be right now
14:33
and $5 the table of contents for online help so the full a toolkit documentation
14:48
but take a look
14:51
documentation okay oh you we have prerequisites overall for working with a toolkit and then you got a student for every app processor let's look
15:18
what notes are generated results elemental results wishing file I'll be talking about that briefly but this detailed description of the day of the definition file if you get advanced with essay toolkit you'll probably wanting to manually edit the this definition file because it's faster in some case
15:45
so documentation totally what's generated by a toolkit in terms of result in what format and a bit of theoretical balance of what was the how to prepare how to prepare with files well your NASA kind input file
16:09
let's help to the table of contents centers on context help so whenever you're in the dialog you just see how now for example go to you'll see how here and then the next help will appear for that of that dialog
16:30
what's new so relatively maintained as a toolkit and you can see which wedding hands and bug fixes are we're introduced in each in your vision so what's been introduced in 602 - and of course we make sure we support the latest versions of NX and MSDN Astra so we talk to each one and then we click that so 601 6 and also using I feature if your version supports so this enhances or that you can refer that page here
17:15
and here's where I they need to rewrite because these examples are based on C maps so as you're starting from fee map and there they're dated so I rate them or we will rewrite them and provide you with with example I mean you can there step by step so you can certainly go off and do them but some are you know it is my week to fee map or not relevant for you guys
17:48
troubleshooter
17:53
you know the typical typical pretty we can't open a no fee to file something happens when you when you open it typical thing are you sure that your master and solutions finished make sure your there's no fatal error and your nasty on that that's one of the common things
18:18
come sing as well I can't get my the results that I want an essay toolkit while it's a process so you have in your price Tran solution your Salwa know if you have to ask for the results that you know you're going to want to process so that's typical gotten to ask for stresses say and then you want to in este toolkit you ask for stresses nothing comes out that because they're not in the op2 file
18:45
black rebels and now you have to you have to put you have to set depressions in Excel to run macros except
18:59
and for you guys is knowing where the Nast ran and the ex help is so there's the documentation for an x10 is available of course you can download it from GTECH you can also a reference it on the internet you don't you don't really need to download it and what's about the NX 10 I believe this is the version you're using and x10 is that correct
19:28
get on it five okay those of you that are around ten even if you're on a five you can you can do this I can see this side uses it's built on right again which is which is a note but what's nice about it is it doesn't run Java applets so it's not always a it's a little Java problem Courage's is better than previous versions and search really works when you put two resigned vectors here you will feel you will find the hits that you want
20:04
so this is a documentation and if you dig a bit into an X I have an X and a strand documentation here and one thing that's really handy is the quick reference guide the quick rise got a description of all the keywords of the cards and some of which we'll be looking at today so if you or what they mean this is this would be where you'd be going here
20:38
you
20:40
there's a little start button on the bottom right so find your once you've picked your op2 file you're in for your groups you've clicking tools you've defined the contents of your processor your yours that you click the start button and then where we actually launch the executable for the for the core code the processor and it we did the hotels fix the presser and run a definition file then tells the processor where to find the definition file so
21:20
it's kind of like an a you know winning on XDA and you want to do an efference all well and he's going to write an input file faster onto disk it's going to launch an Astra NEX acute Abul until next gen where to find the the the national foot file so the similar process here
21:41
you look as you see here in them and the monitor are also stored to a log file and to run into a problem it's good that you you want to you want to debug help you debug the problem then you an email the support at my HTT comm attach your log files give us an idea of what message what the problem could be
22:09
stepping at what is s8 we'll give done it well it's funny thing it's a no filereader so we don't mess NF solutions we return file directly and that gives us a lot attune in terms of being performant and getting what we want what made us a toolkit performing reader is this quick you can use d map or if you don't have a D map like and you next Astra and you can point the delivery databases
22:45
this would use this particular to get to use the residual flexibility method to account for more truncation well it's a youth and I've been showing you to the yearn effects and that's inset kxe in the installation directory uh triple is called statue exe and then I write I specify particular is that you can run satchels dot exe alone you don't need to start the executable fat tools on any mission that's network that can query the license and because we recognize the importance
23:29
of you to be able to run in bash so you can you can actually create batch file that I'll be doing that for you today we have many obstacle to do to freeze your your your desktop and to extract maximum performance because these jobs can take time even if they're faster than our competitors and that sa toolkit also if you want a report writer we dated that HTML pages we populate Excel files and we have macros that operate on the Excel files to give you a report ready type of results
24:19
okay session 5 else as I mentioned before the matrix is going to create a definition file determination is a dot InP it's going to disk and invention is a project underscore the processor iymp can do is you can manage manually edit these theft files or once you what we rated you can reload them into the entities the interface if you're going to process or standalone it means you've created one of these files and
25:03
you know you're going to reference it and I'll show you how to do that
25:11
so let's get a definition filing here
25:22
I've missed some I say toolkit runs done already
25:31
I'm here so this is a rep open to file for it and people you can see is cut the under strand your name so that's what we know what it is so this is synthesis all described in jail in the online help I start meditate you are going to find that you know it be very careful about how you do it you can't introduce just any any keyword anywhere you you know this is that the processor is assuming
26:10
so one of the three locates this is a load case one this is for the random processor so I'm loading an axe I've got my PSD defined here then I got a second load case and it all comes from here a number of load cases here theme equal to three so what is expecting a second and a third and then he has D definition for the second PSD definition for the third the way the PhDs are written is frequency PhD start and then at some point a PSD
26:54
which end here with the PSD is broken into three rules here so I modify this so for example I could say well it's going to set last last row away I'm just going to put two and then we are going to go to directly the 2000 Hertz and then we're going to change the level 20 squared for her
27:23
and I could just
27:27
where you don't have ages here I normally do it in the other dialogue actually myself edited the input file and and you could you can read the input file and then edit it in the dialog I want to show you is that you edited by hey that's just the bottom in especially since you're you're able to run a statue Exe off the command line and II need to manually edit these files can be interesting for you so well you can remove a load J you know again
27:58
if you only one location get that one or two right so let's let it open on this file here
28:15
is that the my file that I've been processing here with us in InP file so I can do is in a random analysis here I can just go point to that definition file that I just edited right now full load so so now that is that I've made here are I can see why yeah I've changed my PhD here to 0.3 I go to tails and hurt yeah just go okay adding here and that lick start and my processor starts tricky
28:56
though is you don't have to load it into the IMC it's just into the the random analysis dialogue here it edit this puckish and add whatever I want here I can add a third row say what is research say right then I can just Brett okay idolo put it this part is that every parsing input file so you got to be careful when you do that I do wrong here oh I forgot the plan okay how it's taking it so this isn't you have or if you're going to be a high ends are safe they're going
29:57
to use this this tool quite a bit this might be interesting for you and the rent my presentation I'll be showing you how to use the the dogs in the user interface just so you know you can do this
30:15
so I mentioned you can launch this a toolkit by clicking the start but I've just done you can start it from the command line and this is as follows so you you you point to this tutorial you got four arguments the first they don't be to filename this one is a definition file name so seeing what the what default syntax for the the definition file is and how you can find a project you can change it here is where you cannot define a new project name this would
30:49
be the third argument and deal entered your code since Assessors we have one executable so the interaction points to the processor that you want to run I'm going to do that if you're going to let manually you have to make that in in tool and then the show because you can do this on Linux as well here I mean the syntax I'm showing here is is for is for widows is a Windows command tool but it could be run off of a linux shell you need to device
31:25
and server the location of the installation I'd like to UM do this right so yes multiple don't need a batch file here so I've already gone what didn't for you guys running it off the command line
31:59
shurl knowledge is good now much okay okay you can do pretty well I'll just do this editor and I'm just dumping the contents of my slide here I need so this is environmental are just define a location yeah so I say tools dot exe is one level down here in the installation so if you go down you're actually your side Tools installation
33:06
you go 264 there there's the core suppressor executable right there so things faster I'm just going to paste this line here
33:32
path to my executable my file name my condition flame and you name I just test like trails February 29 and their crowd this came running in Bowl summary so it is so date here's Bell
34:05
so this is a list of what you see in the is a log file basically so out on the processor here
34:31
good
34:36
type of butterflies a toolkit for a produce well talk about the log file so the log file has run messages and warnings and error messages is that some file it's a summary file it's really an echo of what of the setting we relate whoops we round text output as well so it's kind of bone most of the time you want to put some the day directly to EDX or but sometimes it's nice to have an output file twelve to two pars and we know this and the random processors will
35:23
create HTML pages with pictures and PNG format with dreams so could be quickly the double use for quick grab reports we have an excel we create excel files and when we created the toolkit we we could for both Linux and extend the window so you create accept files even running on Linux by this via some of the libraries we use we can any files that are bigger than seven megabytes which is pretty limited so dump all your data to an excel file you probably want to dump our
36:03
group data or summary data once the file exceeds seven megabytes you can't do anything with it we dump all the contents of the X also to an XML file which does not have any file any size limitation so you have a you know 2gig of personnel file if you want mostly though for you guys is write data to OB to file for the processor Dopey's transform at it is binary to you a universal files and personal files and this is why tax is going to allow a post process for both your
36:45
contour results while data or for your XY plots your chart so if you want results as a function of frequency for example that you know free power spectral density of a accession or you or your stress composite function of frequency then you can use one of those for possession at the beginning of my presentation we call the dynamic tools so the sine and random events that we support our bass-driven okay the vertical application of bass-driven vibration which is
37:31
prevalent in several industries automotive space I guess eyes and what's up when you're using that type of exciting it is you've a resource mortal solutions we're not using any direct it's going usually I can hear some ground noise so what I want to understand your mortal basis because you be specifically some modes and not perhaps not others in your modes form a valid basis for in terms of give mass do we have enough rigid body mass included in all the mode here or do we have
38:18
mortal truncation effects and is that effective mass as well so a processor allows you to do that and the energy processor helps you to understand where your structure participate I'm any code so if you know if you're analyzing a single part it's not a big deal but analyzing and assembly and you know you said which part of the assemblies is responding or which part of the assembly is is
38:47
being trained as a function of the response then the energy processor can be very interesting so let me key now you gonna lies single parts are you guys analyze assemblies typically
39:01
always
39:11
so in this in this case you may be the energy processor may be quite interesting for you so one of it is and the energy like like all the other processors it runs off of an a strand solution in this case a solution 103 may plain type of the same op2 Flex relationship we use for random as you know post process under your model strata kenner G is for for selected groups or for selected load and there's an output that's particularly useful we retros and we generate the
39:54
type of automatically generate the type of plot you see here Aaron Burr and your user your kinetic energies is function of mode or group so really I have an understanding of yours in your assembly
40:13
so like it gives you a measure of which masses witches are excited the most mode so let's combine multiple on the elemental mass matrix by the velocity eigen vectors
40:31
sense Tran requests so then we'll generate these kinetic energies for each of the modes energies so strategy ops you determine which for your structure is experiencing some strain and puted by multiplying amento stiffness matrix by the displacement eigenvectors
40:57
um someone is elastic potential energy or energy of different vaporization and when you when you ask for scripture you'll also see Nastran will also compute strain energy density results and rehearing it's dividing by the name of the element and expected I understand that Islam
41:22
so about nast Ranson tax before so this is the Nastran syntax looks for your solution 103 puts in the energy processor so of course you want to call a solution 103 normal mode solutions then in the control this is a select your output requests and your line labels for loads and boundary conditions you want a pity strainer en DK kinetic energy final here the reason you won't find a threshold for a strain energy is that master and will
42:08
not dubs strain energies into the oil B to file for heritage reasons so be missing some strategy unless you define a very low threshold say zero or very small number
42:21
you can there I'm here so in the bulk data you can use this Pam Comet I need to define the threshold for strain energy
42:33
then that next a strand solution under the parameters tab you can edit the solution for us and of course under the case control tab you can define your output requests now by the way I will send you a PDF of this presentation so the records this might be interesting I have one optional here in terms of the of the set for each processor is in blue here so the param tiny is optional
43:12
nastran solutions from annex master and from from sorry from for my neck
43:28
okay so I'm showing you then I'll be I'll be showing you add up that up for each processor
43:41
and that does suppress strategies by default which have which are small I mean and you want to really what you want to add them here because imagine a global mode imagine a a beam engine that you're modeling a beam using solid element for some reason and you're getting many solid elements to model that beam well I think energies they're all going to be pretty small for each each element but yet together they find the first mode of that beam and what
44:17
you know you need you need capture energies in your open to file to understand the Mo's so the answer dialog looks like this any study similar actually it's at the top of the dialog you can um finition file and load it so if you create a path and you want to use the group depth hands or what else you define it you can just load it on you strain energies or kinetic energies is both and default is you use all the elements but what's more interesting I think is
44:56
if you have split up your model into groups and have a processor tabulate this training kinetic energies by group even if it's comprised of totality of your model and that you can pick the natural frequencies that that you want to contain your processor other by default it's all you can click mode frequency ranges so if you don't want almost first question we write data out to a text file or ASCII file and you'll be Excel here so I don't know why it's not on my
45:33
default click Excel and it will go there you know by default the data written to excel and the text file or for the for the group's only that you pick which you can also the listing of all the elements if you select mints as well but then your your file will get much bigger
46:00
so yes and this is this would be a text file on foot um you know for my mode huge ID the frequency the name of the group and this is an example we've picked a variety of groups here I need the percent strategy for each each group then the total energy right which depends on in your group
46:33
out all the of them then your total may not be 100 total is a total energy for all the elements in the model for this of the of the groups you've defined and she 4% I won't be if you your shoulder haven't been defined
46:55
by mode
47:01
in Excel format where we have a matrix where the columns are load up than the the group signed by Road and and then who is available so we have a macro that's available when you use the dropship I'll do the sort by group or by mode in your excel file and I'll be showing you how that works it's a nice 3d guff here energy by mode by my group number or group name rather which mail port
47:45
so what processor here any questions our scissor before I or why set it up
47:59
all right so what I is a model of a space antenna actually so it's a positive structure like feed here in the middle
48:15
metallic deposit in disconnecting rigid elements or coincident nodes so it's really small for the purposes of my demo we're
48:26
red feminine a strand Simmons and I'm going to do is create a new solution
48:34
solution values
48:39
I'm going I'll be new solution for each each of these processes I want to know Tom you generate you can create the same op2 file and use them in all processor where is it this is between them but I want to show you what they are
48:56
so for create I'm gonna create this pair tiny air to leave us so I'm like click on trail objects
49:09
and the supports all the parents that nast ran annex master and the version of which to support so I go to s to t need something small to put 0s Bo's turn
49:37
edit I'm going to create a new modeling object for the output request
49:48
an energy
49:53
there's going to be defined here for straight energy
49:58
there Jay
50:03
dresses are generated I don't need them I can turn that on
50:14
ah I think that's pretty much it
50:20
by default the first headed and
50:28
that is is this constraint here set a single constraint at the base of my structure this is important if you're using excitation you're really exciting a single node and connecting that note rigidly to the interface points on your structure Platinum's I mean it's just straight brother this or before they're different I saw these come the shaker essentially yeah you know this is this what this is what's moving them down or left and right
51:11
so I'll just turn off of
51:27
so so I have a question we do is a said to look it but you know you just say that you want to have a single note so we'll define but let's say you want to you want to extract 1 Sigma surrender analysis reaction forces at at the mounting points rather than just a single load if I understand because kind of difficult to to get that at those mounting points so no use I mean what you can do here is you can adjust you can put up I mean you could email it if you want what we
52:11
said video this is we we will put bullets elements is what we typically do between scissor down on it and it's true where there's an interface bolt here and we'll see from the from the random processor extract the forces for those Springs those could be beam elements or would be MPC forces so if you ask for MPC for stuff but you will get force of these nodes here you could use those
52:43
okay I'm getting my first ten modes are between 30 Hertz and 59 Hertz and even though you can see this laser is the strain energy and the kinetic energy that can be processed by an X tension the density results as well come
53:09
so this means we have an Opie to file that's written today and so we're going to go process it so
53:19
I'm going to do it all
53:26
Yop pile crippled kinetic energy strain energy
53:38
force
53:42
and so I'm going to export these groups from this model so if I click them
53:52
up what five groups here
53:59
my mom groups for some of the panel's here rubber flap
54:11
and you want right click and go to raishin and let me get my frickin window here so you can see how ninety four elephants at fifteen that I have one mesh pretty nice thing about meshes before let's let's look at this group here I did on purpose this feed rope here my donation on it I four nodes that I put manually I have one match but if you go you know automatic group for example and you go which is what a day you create groups with the same physical
54:48
property it's going to dump meshes in here so you want to be able to export the unisol file format you need to have elements so what you need to go add to group you see this as element pick these elements so that three hundred four I go okay then we'll pick group when I go information
55:17
I have but the discrete elements so this is what you need
55:25
as I'm going to Universal file so I can reuse by the processor I'm going to change I'm going to edit my fam and I can check any tennis all right I'm gonna pick ideas U and V here the personal file format that were speaking
55:46
so now I'm Universal interval language or an export it and go file export simulation
55:59
yes I want and then define an output file
56:12
even me yes it is for everything you know this mess of physical properties elements and then its type groups export it zero goosal group skipped so don't want to see here if you only have a group with a mesh in it is going to say one group skip for example fuller to etc
56:37
you
56:43
good night I like to EDX Natura for them and do and toolkit a main dialogue I go groups import from Universal file this is the fudges Gordon here
57:07
so I did it but you'll be able to see the groups later in the tool so I go to tools Don did I have I have energy analysis it's available to me and we're going to pick all the groups that I've we've read in some groups groups developments in them switch them to inactive window to the active window
57:42
like reflect our main reflector feed and I'll do the same for kinetic energy and you can I mean this is something you can do you can just create give it whatever name you want you can edit label so upset good 100,000 with them interviews as well using these these brackets so I can go from in two hundred
58:20
increasing ah I don't this case I'm not going to use it I'm going to press all the modes and I'm going to dump my data to excel things so I could I log and then that writes the definition file and when I click start the tool processes the definition file with the right processor so it's run I got no warnings no errors
59:00
you
59:05
but uh I got an epic for to ask you file
59:13
by the kinetic energy results for each group
59:20
more interesting file so I'll open the Connecticut the connecticut refile it's in my other screen your home
59:32
hey ro you can check my rigor to whatever you want template file look at the river so what are my mode what these are my groups I'm tiny my 100% so reading and my grew don't overlap so meaning that the total energy in each group if I add them all up I get a hundred percent of the strain energy in the model which is their phone here what is no formatting say
1:00:19
so so for the Lecter unenrolled for the ones with lots of kinetic energy or lots of this placement for the side panels is modes 2 & 3 5 6 8 9 10 for the server Lecter is mode 7 & 8
1:00:40
what I can do is put photos and then I can sort these by by a group say so I can say let's look at there's nothing but for the feed let's say the main reflector so it did the groups by descending strain energy so group fours got the most strain energy for this group and one is it got sixty percent can pick another another group and then I'm descending a strain energy
1:01:21
like this and then I can sort by
1:01:27
it's like good one let's look at let's look at by descending strategy which group is more important so the main reflector has got most strategy than the side panels
1:01:40
these complexes can be really valuable because it's it's scale where where stuff is happening I mean if you're it might be obvious when you look at an animation and which part of the structure is straining certainly you might get a feel for what's displacing in a Kinect for the strain it might not be not obvious that's generated for you so we have a 3d chart if I click here 3d rotation
1:02:13
this is ability here
1:02:22
once I'm a pro I could just save it no matter the next time I open it
1:02:32
for energy
1:02:39
or kinetic energy
1:02:45
you
1:02:54
Oh
1:03:00
you
1:03:07
for kinetic
1:03:12
of course that is completely different because here were looking at the displacement rather than strain or velocity rather than stress
1:03:27
is this something that we guys
1:03:33
please okay
1:03:47
again this data is in and then X but to get it you know you have to format it quite a bit so to be able to size it in the way we've shown it
1:04:01
talk about moral processor or the model summary then which modes of high effective mass and high response actually as well
1:04:14
so this too is mobile effective masses or weights addition so the percentage of modal compared to this is body weight any direction gives you which are the key mode so if you lots of mass participating in a mode that we have lots of stress when you excite it
1:04:42
also look at Rick Sharon what we we we create a few harmonic excitation at the base and we allow you to pick or sorry know the degrees of freedom at which you want to see responses go to identify local mode you know they have small effective mass but still have high responses
1:05:08
so lots of mass and moat lotsa and then also modes with lots of response and together gives you a pretty last picture or your problems could be on your struct which moles are important for you
1:05:28
and but wait is the specificity of the modal participation factor
1:05:38
attraction if your is your mode effective weight weight very divided by little components of the mass matrix here which Nastran outputs to the grid point weight generator that output the grip on weight generator in my next
1:05:59
and factors is a prop that generalized mass matrix so displacing is vectors the reduced mass matrix and the rigid body modes
1:06:11
yes and stop messin wait but we wait here because this is wait master that you can put it on so there's there's a per Internet friend called wait man that's essentially master wait if you want and you can use it to help it really scales the mass matrix but in the preprocessor allows you to the per unit of mass and all mass related properties in as you wait so for example you can define spam
1:06:47
masses rather than pound second squared French there's this great way generator she the weight the the not nice so if you've got a weight mass crate it's accounted for in this and just table
1:07:07
what you newark and typically
1:07:15
for me back and forth
1:07:17
depending what mine NX will allow you to find your your your mass is that if you're using and using imperial units you know PES I allow you to find your masses and found masses and does all this for you but if ever you run into a model with weight mass parameter in it which which does similar does the same thing then this gives you a bit of a background as to why it exists and where it's used
1:07:51
and this is a tabular or ASCII format we'll give you some your eigen values your modes princey your a sort of mass by my coordinate system X Y Z Z sorry and then your fractions also add the rotational degrees of freedom yes in each direction and then your cumulative effect of that and this is born because when you base excitation mobile pieces should include all your you know when you live as a total infraction there should be as close to a hundred percent
1:08:31
of your rigid body mass as possible in each direction and how are from that her percent tell you have to use we have to compensate for modal truncation some truncation is your missing rigid body mask really in your model basis is what it is and and and we'll be discussing that later
1:09:00
so that was that was for effective and you know going here you can NASA right to this out to the fo6 file you can here you can edit the Apple six file of master and and just copy and paste it into a spreadsheet if you want I personally don't really care for copying and pasting on a recurring basis so much prefer to use this this form is cool so what those are also down as I mentioned before is it calculates a response to a steady-state excitation of 1g every time you
1:09:36
specify so what does it well if you look at your point of motion here you you bet you express it and in the model based just so you express placement by a function of the eigenvector moment is that is your Q as a function of harmonic excitation so a format ae2 J Omega T and to get the harmonic respond say bond that we calculate in total summary and so in the sine processor which which is the next section we're going to look at right is your seller ation plus your dynamic
1:10:19
response which is omega squared times the sum of the transfer function H times the eigenvector
1:10:32
at times your emotive participation factor and this transfer function H is one is classic here you you must have seen this before is one divided by Omega squared so I'll give piers here as your is your front of frequency Omega is your excitation frequency and then plus J Omega Omega P divided by Q for that application factor you are two natural frequency the more responding it and the more amplification you have the higher it's going to be your results that is and
1:11:18
you're probably huh beside a processor in the Nastran deck using master and syntax again using a solution 103 normals and Alice's either control
1:11:34
and they kick you requests placements and it's optional but if you get a post process responses you've got to put displacement for at least a set of your model or all strain your model require rest mode effective mass from nasstrac and you can send to the op2 file so it ought to be here with the option plot
1:12:06
a option here is is not necessary you can specify a grid around which the rotational are account for the account work for and the BND parent post - - and that's the edx default so I next I mean it it's always going to write parent post - - it's always going to write a current post the XDS so I'm just put formation and that's what a strand next outputs by default now if your mass or weight frictions you're going to need to put this from here gr DN T comma X is
1:12:42
a is either a node or if you put 0 let's the of it's the origin of the global coordinate system not greater and that's optional but if you want for actions I would think that's the 10 you to put it in then USBC card so your your no constraint at the emotional processor Doug at the top as for every processor we have the look the name and path to the definition file protection is a section for effective masses in section for a response prediction
1:13:22
choice with just translational masses or all course where you have rotational effectively classicist which will be can I imagine that can leave or be then you get a high rotation rotational effective mass for the axis that's out of the side of the paper move the screen bigger so a percent which is the sin of the rigid box so any mood with any component that has more than this percent effective mass will get hided will get for response you
1:14:01
can select groups and there is to be shown by group second elevation method is being either exact or proximate if you use exact method we all the modes in your solution 103 to calculate the response if it's approximate we only use the mode that corresponds to the excitation frequency
1:14:26
sir and a data filter so the new filter filters out small responses so I feel super response is less than so many gs agha less the motor it means that I want to see modes that have more than so many Jesus response for the node that I've selected x-height result combinations off amplification factor and your excel file toggle which truly should always be on and I believe it is in this dialog we're going to see for the response Britain's it like this you're going
1:15:03
to see case ID and then you're going to see to locate case G X 1 G I think of the response for the nodes you selected by frequency you're going to see the X result X magnitude the Y magnitude the Z magnitude and then the root sum squared magnitude of all the components and then the the magnitude here one of the components has exceeded the filter you've defined so it was 15 GS well here I have 18 20 and so this is what I said to be a critical mode here
1:15:50
this mode mode here three 4.98 Hey
1:16:02
and you get the effective mass results sorry then you have my mode frequency uh your mass is here in fractions and the responses the maximum response which loads of actual response which note and occurred at which group and if the star is here essentially is asked to show which are the important moments either modes that have lots of effective mass like a 23% respond 3/8 G's this second mode here has got 6/10 effective
1:16:38
mass and X it's got a 43g maximum response at this node so you can ID which will have a lot mass participation and have hot responses both may not hire and you can if you want you can just get them from your analysis or at least unit you worry about the when you post process
1:17:00
while we generate anyway we will create rosanna for you i and the grass weekly are effective mass by degree of freedom by mode cumulative effective mass by the area freedom by mode we have formatted tables so having asterisk we height the row for the modes that meet the criteria the most filters or office and for responses
1:17:42
you
1:17:47
so why don't we demo on how to use that tool so I added this solution that I've created and just modify it a bit I'm
1:18:07
again one of real life
1:18:16
and DNT parameter here
1:18:33
to the basic one system
1:18:45
I think I'm going to put well placement equals all - what else do I need
1:18:54
Kelly this idea
1:19:02
so if I go out put request
1:19:09
model effective measurement check done and I'm going to make sure the medium is
1:19:16
that's important and it doesn't work
1:19:22
request that works the other way then the send all the others disables stress I don't really need it
1:19:33
and then I have to remind my constraint my constraint rank concern go add to a solution intercept
1:19:48
I'm going to solution
1:19:52
I'm using a ten point two and NMU next ten point oh three but all this will show in earlier versions makes my results
1:20:07
again to say
1:20:11
I'm going to do going to be to file I just jet here and you can see - so you know now gave my solution a name such of it I'm using this
1:20:29
what this is to use it for and I'm going to go
1:20:40
I want to see all acids that have translatable map components greater than 10% of the rigid
1:20:51
former unpick
1:20:57
I'm going to clean those and feed here
1:21:07
I'm going to solve them I want all the modes and have more than say 25 G's of response for one GX tation I don't want to see it uploads any responses below 10 G's and I'm going to load in the X
1:21:26
we file the bug settings they've been removing to add a definition file and now I just start I should definitely I will be right now I just start
1:21:45
that's good
1:21:52
which okay the definition files being written then straight to the Excel resolved
1:22:04
and again we have made on sheet
1:22:12
some tab you have a neckl fear of your settings that's charged for effective mass XY and z of professions so if the X fraction isn't right Cinci mode is a is important in what what direction so the whoops why mode is no sticks too much NZ cumulative s you can see it's really due to the X direction almost load some percent Africa ten Mohs
1:22:58
then we have the away
1:23:04
so we effective mass by five by translational degree term and then the fractions follow the next three columns
1:23:19
you can't hear say what singing effectiveness
1:23:29
on my mode back to this this is so I'm uh I had a bunch of nodes in this feed group and I want to see all the modes that give me more than 25 G response and the ad is mode is moated 4700
1:23:49
pick up 35 G's magnitude because you know these are complex result of magnitude is 35 G's and 11 in disease free small our magnitude displacement our acceleration is about 36 geez
1:24:09
in X and for loading into Z I had you know five ten G B's only if I look at em right now so it combines that modal effective mass and the responses that are interesting for me and then show which models are important so which Mo's have either Mort effective mass in any direction or have more IG responds and my group PT feed I could have put in groups here and the group name would be here
1:24:41
look in tonight and one in one here all the modes have put a lot of stress become effective mass portal stress because high response
1:24:59
so it looks pretty good in a report just copy and paste that
1:25:11
okay
1:25:17
you
1:25:19
that's something thing for you guys
1:25:27
here yes but identity for everybody good but it's a pretty way of dealing and I've normally been doing this from the fo6 file just grabbing it from there but this is more or completely gives more information it's personal some people in copying and pasting is just after a while it bugs me so you know I just let the software do it
1:26:01
all right so two processors that you can use the tab and that can help you understand your Moses like what's included excellent usually added the really great animation tools to visualize each mode or the important modes but to processors then you have a animations I think pretty funny standing what's happening in your assembly
1:26:30
so now your sexualisation
1:26:35
so my brother at first and always been using the random that this there's of overlap between both so it's not going to it will be pertinent for you I laugh you ever ever uses processor
1:26:59
today's yeah okay so you have requires that to perform a harmik basic citation analysis I
1:27:16
this could be this could be burning for
1:27:24
so what this or does is steady-state he nicked a transient the response look at state so it's a deterministic periodic signal input and response it's not firmly enough it's get ready to be doing trans now this
1:27:56
so in on his sign you sign test you do a
1:28:07
so what does odd is is it it Center to what annex response sim does it is it basis elf off of an ashtray on normal mode solution directors and grabs the rigid body modes and then the responses of your model a harmonic based acceleration equivalent of a response a ssin and also or Nass trans solution one level except it sits unused and it's faster well design of the random processors they share the same master and solution sequence
1:28:46
share the Santa ranked cards and the same restrictions in terms of which cards you can use
1:28:56
eternal assessor are the ability to include the ten sub cases in a run or a sub k the combination of direction and input spectrum counts for modal truncation so we looked at you know the software the cumulative effect of mass so if it's not as high as you'd like it to be say 90 95 % of your rigid mass and you product turn this on if it is that you can you can use to account on occasion so either residual vectors which is a default method or I'm residual X ability
1:29:34
which is something but differently using using a dmap so this is high-level national and I don't figure with it I don't know if you have a license for it it is an option key place
1:29:57
you
1:30:00
I don't know anybody else use it okay map you need a license however we we said we ever work around such that you can use the the tape ability wind up without needing a licenses so I'm going to discuss that briefly addition can be done using all of the next Rance computer or you can select which one you will use and it computes the absolute or the relative displacements velocities and acceleration so relative be interesting because as you know it
1:30:37
if you're under guys go to the go see the end when they actually test this cuz I'm testing a random testing typically done on a shaker the shaker itself moves something a relative will remove at that displacement of the excel of the input no J curve which comport inton for those mix so you the Maxima over all frequencies so my maximum response over all frequencies it'll sit by a vitamin C or by a by result which we know elemental which Tran world is known
1:31:25
as sort one and sort two options if you look at the at your control output request that's how they describe them warriors of the assigned processor is that your results are complex yeah so you get calls for everything so um what's for me stress we figure out the phase at which the stress is met amusing analytical approach which is very fast and so we'll get some real one may see stress and we will compute margins of safety based on the von mises stress automatic
1:32:10
you we generate XY plots which you can HTML pages in Excel or in the edx function navigator are you guys familiar with the edx function navigator
1:32:29
okay I guess I'll be giving you a little introductions that today
1:32:41
so vector approach I'm forming model truncation we call it the MF mass approach and I'll add to an aim is to have this effective mass request in your case control I shall for for the model summary but single here for the sign and the random processors unique of course a constraint of the single load we've talked about that and you need all your requests so if your requests been velocity or acceleration from the toolkit you need this for displacements
1:33:20
in the nationís all 103 solution if you're going to stress is from the toolkit aspera stresses of the solve 103 solution similarly for any other output requests like force MPC bars etc that's forgetting to add the request from here in the bulk data for in from that I put the post this to and post the XDS um well as the SPC card to use the residual vector approach which is the MF mass approach you need to add param Rose vine er yes which you can using the
1:34:01
modeling objects as I've shown before and we're going to do it here good Bible if using image it's a bit different you don't need the MF mass card but you need to point to the D map use a statement or if you don't have a d map license then you point to a delivery database plane if you go in the as you guys si - look it's a lot help and you find it doesn't matter to be much
1:34:37
it to you what you need what extra car need what params you need how to out of the D map or to a delivery database at the rear database essentially a solution with that includes the D map already for you so you don't need to invoke the map
1:35:00
a PDF of this so you'll have reference do you need on the o ye om a parameter need to be set to global like it doesn't remain them for displacement yeah I didn't add it but no it doesn't actually happens because we're not writing the same data block as in random but that's question
1:35:33
is a pretty huge Ecore global so if I fly step ahead a couple of slides here and I go to the random processor
1:35:48
I'll get there eventually
1:36:01
okay forget that damn you G chord you need to you need to set to toggle on or is it you're a for acceleration displacement or velocity maybe you need to do that here
1:36:23
the reason is because as I put notes here and the op2 file doesn't have a data block that's in the basic coordinate system just for the dolt but that's but only if we're not but we're not dumping data to the ability to file an assigned processor so that's why it's not required you
1:36:51
story so you'll need it
1:36:57
all right I'm truncation okay this is from the annex master and basic dynamic analysis you guys know I do we're accessing the NX and the index master online help that guide can be found on there so we'll actually it will happen that not words are are computed and certainly possible that you compute too many modes but not generally the case but it's never an recommend that you should involve the modes whose frequencies lie within the range or the
1:37:38
forcing frequencies as a minimum if you're if ever I feel so you find 100 Hertz then you should at least compute all the modes to on earth as a minimum and for better accuracy this guide recommend that you modes up to at least two or three times the highest forcing frequency so you're going to go 200 Hertz and you should pick 200 Hertz to compute all the most the 300 which overkill but that's really covering all the the covering basic vectors can also be requested
1:38:18
to improve the respond by adding the missing static Leki associated with the higher modes
1:38:30
vector approach then against the same basic analysis use dynamic analysis use guys
1:38:41
perfect the only because upset of all the modes for the structure I mean it may test for performance reasons so all computer modes may represent most of the dynamic content of the structure of the amino fire frequency modes can lead to errors resulting response prediction at higher frequencies do contribute statically to the response and to a pallet problem is to use residual vectors day-olds vectors do contain a set missing static
1:39:23
so again choose the first one is the is the residual vector approach I've just talked about here or what we call the MF mass approach
1:39:36
and here calculating extra eigenvectors right from this reference here looking at the total forces as as and what we're doing is we're removing from the total in forces so the forces you get bike is moving a moving model this b and you subtract the the forces the total forces that you subtract the the component of the forces from the that solution
1:40:23
okay then using the universe of the stiffness matrix cat transforming these forces to displacement
1:40:33
those are the same
1:40:39
again this is the tribution and this is the contribution from the Moe's you retain
1:40:48
those vectors are orthogonal with respect to the mass and stiffness matrices there under the existing mode shapes whether your actual modes you request the ones you're used to getting to be a residual mode
1:41:05
and that mass difference makes are updated using all these most the total bases here and represent the static response it's orthogonal to the restrain mo to the retain mode and they can it in the in the vectors that you're using for
1:41:29
per position
1:41:36
the map approach or residual flexibility is similar so you'll see that the requires came out to be very similar if you look at the excitation frequency is lower than the frequency of the highest retained mode where our means for retain okay response says this is Z is the total response as the Dyke response plus the static contribution the key contribution of the missing modes which is that the static contribution using the static using the stiffness matrix
1:42:11
slightly for you to enforce acceleration the contrition from the retained mode similar to in that second the last slide the difference the main difference is that this cushion the static contribution is added to the responses directly the right extra you have to manage which is nice so you don't excite those modes they don't exist they're included already in the response
1:42:46
uh next response sim you have the you have a similar concept called the model acceleration method it's just similar which which which sums up the static contribution of the retained modes and then subtracted from the total static contributions
1:43:08
so how does motor truncation show up in your results um I mean for some slicers you'll see the difference really being between the natural frequencies so actual frequencies you get such a huge imaginary response that really that the dynamic imaginary responses is is basically everything between owed to see something different with this flows is the result of for now we'll see two nineties the solution 111 which is model and solution 108
1:43:52
which is a direct direct solution doesn't work in the basis and doesn't so it's not subject to multiplication and that's the really the exact response is it with the X's here here so you can see that off resonance we're under predicting the response wishin 111 with now to win of that residual flexibility we tend to follow each one of these solutions so the width or flexibility this dashed line follows the Nastran solution went away pretty clearly
1:44:32
could of a direct analysis in the model domain with extra residual flexibility components the less Q effective mass you have or this is important if you know you do a signing out office and you excite off residence you say I want to know what my responses at first and the rent is not equal to mass times your acceleration you know why right you're really effective last year that make sense
1:45:16
so the sign is neatly organized
1:45:22
ah I guess that sir
1:45:30
you're able loading duction file then you create cases duties that you want including your pro super position can you lift amping quick references and for calculation and margin of safety calculation
1:46:03
okay petition 500 cases or sub cases each something to a different output request
1:46:16
right
1:46:19
and give it a flexibility
1:46:24
by default according system but can pick a nomination
1:46:32
the excitation and you define integration acceleration factor applies to the acceleration here so can consistent units we'll talk about those later per second squared but the function in G is here
1:47:00
386 so I'm a factor between G's and is per second squared
1:47:10
so you can order units at this point structural analysis toolkit is unitless okay Tran that's trying to cure what units it's using four units and there is a Chris information message at the top but that's really just for the post processor and that's right doesn't give a hoot what you're using sub consistent units ibex strata stable from P map one training you can see it on that this type of table shows you what consistent units are
1:47:54
I define your Infinix ratio and you find it in consistent units so that's anything inches then he's got to be inch per second square using millimeters x squared
1:48:13
you kissed resolution but the Nastran solution you are those of your sim compound ensure millimeter millimeter is a million you know you know you can select any unit system for enas resolution so I go to my
1:48:52
Texas slowing down the applications there
1:49:00
big time
1:49:10
all right
1:49:13
you
1:49:18
what do what happened
1:49:24
you
1:49:28
next year
1:49:33
you
1:49:39
continue
1:49:44
now solution you can you can the exporter basing your IP file in any unit from your pick
1:49:52
you now the files from the se tool kit includes the Nexus companion results and Batman in the swing units and you're all set
1:50:16
I know if you're from us
1:50:24
you
1:50:27
all right you have a student you can add Mannion results which are special from A to B generally from I see it essentially locked in there essentially lumped into the the main solution well when you attach those make sure you're attached the same solution that you've used that you're one of saw one free
1:51:00
all right back to our output request request need a frequency set which is an equivalent of Astra and bulk data fret cards fret 1 4 2 4 3 excitation frequencies are considered if you're going to have create loss your more frequencies to have a nice smooth curve in but looking with some results you'll probably focus on the natural frequencies so another didn't fly so then this is for tabular output creating
1:51:39
contours and an axe what's plot HTML pages or and plots in India next function navigator
1:51:53
you
1:51:55
von mises 854 10x on the neutral talk bill is on so neutral actually sorry but it's not clear but it generates a universal well
1:52:10
since otherwise are very similar to the random process
1:52:17
you're all your stresses for example here you can select groups and then click the yellow arrow here and that gives you less what's nice is you can ask him result more than once which you can't do in that afternoon so you can say you know you can add for it ask for accelerations to be sent text file and then you can ask for another request for accelerations and send plots say to excel whether their frequency set
1:52:49
the idea is to defraud if you yes when you have accelerations is a scale factor than fear is to be able to post process results in G's rather than inch per seconds from millimeters per second squared B will bring this back into an X we define here so in slot sorry a contour this consistent unit so even if you find a scale factor here we're still change per second squared or millimeters per second squared and
1:53:28
you with in the FedEx function plots the units that are shown are actually incorrect and all these issues will be fixed in the next room please add up then you can have this slide for reference
1:53:44
now the fiber equals is that frequent is where you want excite what frequencies you want to excite your structure you can so your earn your lower frequency as well natural frequencies as well as intermediate frequencies between the band and the natural frequencies using perfect to refresh reformat freeware split up domain and however ranges whereas prek-3 will splits up the quincies between each mode shape
1:54:22
and then you can divisional frequencies so if you've got a breakpoint in your curve and then frequency and add it here your plants will actually show that break point
1:54:35
pretty thorough sees and so that finding your frequency set for and your others the model basis you can fill modes as our mode ranges or frequency ranges if you flex ability approach or the map approach you can check this on or off for the reflux ability effects and then you have a toggle for absolute or relative distance velocity acceleration
1:55:10
is modal damping slide as an amplification factor Q which is e to the inverse of structural damping is equal to 1 over 2 Zeta where Zeta is your critical damping ratio or your Epping motor the only type of damping that the toolkit
1:55:35
find dampening interrupt the model in the material arts or in your spring elements we will ignore that
1:55:44
these are linear between each producing
1:55:52
let's you know we're we have complex results so how do you want then you want real imaginary you want magnitude phase or magnitude only for the tag up mainly and some of the plot I'll detail hi Milo you want group please for the groups or you want or do you want complete daily any to have defined in all the contents of the group so you lost the data you want to sort one or two formats so by frequency part by no to know
1:56:39
you
1:56:41
here's a compute phase consists of on me see stresses
1:56:48
using an ethical expressions on your Miracle Method dentists performance
1:57:01
so you need to get the vos fesses in the bars of safety yet if you just ask for stresses and the output requests it won't delay the one the face consistent von Mises track domination of alle feather all or selected groups a double strength safety factor we use this equation here in generate margin of safety and up stress is it is that you there's no reference Opie to file in the middle for those you define in
1:57:40
annex well unfortunately even though Greg you get rid of an X you have to basically double your allowable stress either yield or ultimate that is just for information
1:57:56
it is still with this for the sign processor is the same as an erratum processor we will look on me see stresses and associated margins ones that are less than the off here by just one percentage all over the place so just make that big make that you know really under ten million
1:58:23
I guess you think in the past it was hard we had a lot of data and sister is tough time listening it not the case anymore your von Mises stresses and your margins of safety over margins that are lower than this cutoff
1:58:43
the Cypress these that all pages
1:58:54
so plenty come out in magnitude format by the way regardless of whether you picked real imaginary for these plots are only magnitudes you can get and that's really nice same plots as an HTML page but you can do stuff with Excel data you for example do caching sad your notching logic and see what they give and you can actually feed that back into the SH don't get as an Austin good you can bring NX because the function tools are
1:59:32
pretty good in actually they're excellent in the next so called a new associated function file at AF you file an or at the universal file that essays will get generated for you
1:59:51
I feel of X output for the Mises stresses margin safety
2:00:03
again irate you need to click button and generate stress cases to get that book
2:00:15
in here current version of the sign processor only supplement central results if you have this on your element only centroidal stresses are processed that will be fixed in the next version v7
2:00:39
I'm so dull tonight next
2:00:46
the sign processor by the way
2:00:56
you
2:01:12
alright let me just ask know why it bailed on me I wasn't doing to it
2:01:26
you
2:01:30
this is going to take a ten-minute break I finish up with a random processor so through a flexibility or residual vectors that's something that you would guys would want to use
2:01:51
you yes that's me but I don't know buddy
2:02:06
you
2:02:12
wondering if we're having lectures here really small
2:02:31
okay well this stuff I'm gonna switch to my lap
2:02:47
it seems like dividual vectors when used from other in a train or whatever any other nest it didn't seem to make that much difference in the result to you Joe Road the only at the look amplitude where seemed to make a difference alright so maybe that maybe that's why I never saw a big difference in like the result because the peaks are the same
2:03:20
and you'd have to know I guess not like that like you showed before yeah foresight for sure that's the case and they all did at the overall Diavel RMS value wasn't with like totally different boy right being by the peaks and the peaks for this thing well this is not just driven by the piece it's also driven by the valleys so if you're from well you're good you're gonna yeah mainly the pink
2:03:49
it C affects the RMS and random and home and so if you have serious truncation you might notice this here with an X try to get
2:04:15
we check my mail
2:04:37
you hundred eighty three hundred and fifty days a year where I formally do do written air for residual effects but I can't have studied it I just hit the bond yeah
2:05:01
well there's some things probably when you're doing like shot response you're not supposed to use I give you the number of an answer yes like Taurus for a first time yeah yes yeah yeah no I think when you do an SRS analysis reason I don't know they told us that in one of her just training things and then I've used and I notice here the results were way too hot so I'll use inversion of it I think our network is crapped out
2:06:09
actually this doesn't start - it'll take breakdown
2:06:32
man
2:06:34
stressful
2:06:46
all right we're sticking break letting a 10-minute break so I'm back at the est days me enough okay washroom break or copy break whatever and I'll try to get a next one here gender all right 30 minutes
2:07:43
you
2:18:56
now lookie there here a table to pop so I'm somebody X isn't working here I'm light it up so project Tendo I'll continue unless you want me to wait a bit more tape
2:19:33
Oh any this you know what I've been presenting is there you know my too much smell is is that comment good kidding that I'm tend to want to concentrate on the dozens to push for the Yardley up now push that button the theory behind is good also is we needed to do no or leads there why are some button pushing soon and you can see my screen there
2:20:18
so what I'll do this line processor
2:20:24
and
2:20:29
again I'll cream on three real eigenvalue solution
2:20:37
you
2:20:46
so the case to find my upper class
2:20:51
we want
2:20:56
something's up and doing real I can buy up real eigenvalues but there are other options also right that does it really matter if you respond sim um it might work for you might work for you the response sim works response team will work is but but the response times a D map or embedded D map will with extra eigenvector is a constraint models - metal molds that in your binary file and so you're going to get a bigger OB to file then you really need and responsive
2:22:00
deem much some limitations or restrictions that the toolkit doesn't have for instance yeah yeah the room seven-a-side processor will be able to work with super elements and I don't believe that response sim does that always it's better if you if you if you start straight from a regular solution one which we as opposed to DM my response simulation
2:22:33
so I try to use the one I'm going to edit this solution here and I'm going to I was doing taste controller on the plus sign
2:22:52
model effective ass we will let plot I'm going to ask for stresses make sure those are on yeah for corn
2:23:12
but my default should be on
2:23:19
parameters and this is where I have an issue with the earlier versions of NX 10 and so I've tried letting me so I say this is sighs and I go to res now we're going to use the residual vector approach so I mask or turn it here raspberry yeah and you see here
2:23:56
there we go
2:24:00
I think modeling object called signing it remember
2:24:09
come res Raynor I've got my output request and I always forget to do right right
2:24:29
my laptop I think will be pretty quick
2:24:41
I should get sixteen miles
2:24:49
oh yeah and my original ten mohs and then funeral directors here phones 11 to 16 you can see the receipt are between 70 to 141 here which I you know I like better
2:25:20
um okay
2:25:31
of the open this time the
2:25:38
on the hotel island forth the universal file after we do this because I switch computer
2:25:51
you
2:25:54
started finding our little cases all system one
2:25:59
Thanks what I'll do is all imports the the spectrum from a file I have on them this
2:26:38
input so what that is is just to it's just two columns that's frequency versus acceleration the internet manually right there instead if you wanted to associate input where I could blow with annual yeah for example I could add up something a fourth year say the show 5g
2:27:05
so you just step
2:27:09
a fiver right here
2:27:15
and these are a G solution is an inch per second squared so I got to put scale factor
2:27:23
I'm in the wrong unit system
2:27:31
I'm doing that as unit will pass what actually it's just remarkable how the stress goes away
2:27:48
so now I go back I clicked on output requests and I'll turn by my frequency set so I'm going to do some two frequency sets one tag or contour data and one plotting the data and we're going to go 50 Hertz
2:28:15
ask for my natural frequencies and nothing
2:28:19
okay and they said in the end that's data rate another one I don't call it plot minutes to it so it gives me nice applause yeah
2:28:39
of course your eigen values I'm going to pick three this three produced arranged between natural frequency right now I get an point and I'm also going to create a an additional freak 225 to where my step was so I can see the step in the block other ones I
2:29:07
got one in my
2:29:16
for margins safety
2:29:23
I'll stick one which is just a financial frequency here I can pick groups or anything I mean it's all defined by the stress cases I will define later Oh sis can that then I'm asked for some two results for acceleration
2:29:41
and I'm enough to a universal file or only for my dad frequencies I'm going to Appalachia no component I'm going to scale factor here quick interesting
2:30:03
let's on reflector
2:30:07
Simo
2:30:14
send out some acceleration for blocks so what I'm doing is I'm gonna turn these guys off and I'm going to have to ask for data in all formats
2:30:27
html5 so I'm gonna skelter
2:30:36
my gruesome
2:30:39
for a response nodes I have a group of smell destruction turn
2:30:51
I've got one load X three output
2:30:58
well I can definition or frequencies I don't think it very well poggle off I'm super the joints ability effect here so I don't think it makes any difference
2:31:13
we're going to keep ABS results
2:31:19
Paul dad been here 20 bucks 225
2:31:28
I believe it's a real imaginary
2:31:35
I'm not going to data in the text while I'm going to look at it and so this doesn't really matter sort waters
2:31:44
where's the margins of safety bays on one meter these stress we're going to create a stress case only the component there is that feeds that horn there so you know all this and we see that handles in the red processor so 20,000 psi hi each factor of 1 point
2:32:14
you
2:32:17
a pickup have a lot of results I'm going to put a high number here for my minute margin max large in the city for prints I want to be able to see car all the results
2:32:31
lambie's for the same elements but for ultimate this time you don't it's different
2:32:41
and which a la ville you're using ultimate or yield so here put a different allowable stress on a different safety factor okay this is only based on that one material if you have an assembly file with multiple materials really the English form arsons safety the way I see it so I just usually put what I am for the bars if you some couple everything like you said earlier yeah you know you just create more cases and then you you prove
2:33:20
yet yeah you define a different material level heinessen doesn't store the ultimate strength and the yield strength of the match card otherwise you can erect the op2 file
2:33:37
yeah thanks out on that
2:33:44
okay let's search this guy
2:33:51
you notice arrests 16 sets being frozen right so
2:33:58
my first resistive 70 Earth's I'm exciting a $50 but side on the mode so those civil vectors are being excited
2:34:19
so let's go look at artery and trip files here
2:34:30
clean because let's look at our MLP
2:34:41
so for a acceleration acceleration plot and I only ask for translation so I don't have only have X going to be here
2:34:54
all right i factor with 386 oh that is scale that's a quick well
2:35:05
I also plots to go to excel so I have this result your call Excel written graph I pros really the singing HTML pages is actually here in the sheet and links to each graph up here and you can see the plur ated save the engine only but at least i i can what is data in excel i can i can do something
2:35:44
you
2:35:47
we're going to be missing two it
2:35:52
when X is gone
2:35:57
later here to file associated function file so I'm gonna create a new walk recall this sign - I'm going to
2:36:19
first off I generated
2:36:31
because I was careful sign I thought you would be this one here
2:36:38
see in navigator all the plots show up here double-click on one
2:36:46
I mean you can then to promote I turn wrong here that's just uh I don't think we write the shock I don't think we write the unit's data set to the usb files it doesn't know what to write in a car so I think if you really fix this you just edit it and it's a bit of a adieu for each one today
2:37:17
with G's here and it it's not gonna not gonna scale anything this is going to change this player I'm gentle as I said you can overlay these things you can have you know
2:37:40
they do things like that when this
2:37:49
I mean you go to log scales really a bundle again already dead
2:38:02
you
2:38:07
Thank You senator responds hello get responsive credit there's some SAP all math over all maps which you can do is you can scale functions up and down you can it envelops envelopes rather and so a bunch of responses and you want to pass on the maximum to somebody while you just back from envelope for all
2:38:37
so very pond and for the random processor as well same thing and show you a bit more functionality you have the red
2:38:47
all right so contour data internet so how do we do that
2:38:55
that's our solution what I do is I'm going to unload it so I've been I'm in the photogate er I pick our solutions I unload right go back to my sim pick that 60 solution you'll add companion result
2:39:15
results okay let me step back a bit in a pose see in the screen in the op2 file or in the results file or whatever access a Bacchus whatever file tree is always written results file and that's what you see the companion file or does not need to have geometry it uses this geometry as in the main solution file
2:39:44
but we're just going to follow general
2:39:49
and what was in the bar the safety they were accelerated
2:39:55
hejdå here so now what I'm going to do is I'm going to go back to the post navigator and I'm going to load solution and now they're the next drop an extra sub case your sign load case one okay
2:40:16
three
2:40:26
my margins of safety
2:40:42
be that much stuff in there
2:40:55
hmm
2:41:00
so look
2:41:03
you
2:41:16
I did something wrong it looks like when it's done right I'm not loaded
2:41:49
that's more like it
2:41:53
I love to screw up in my effect definition John Pennsauken thank you all right so our acceleration and one may see stresses along Mises stress
2:42:07
I'll lead is here
2:42:13
so if I edit my post view I can't convert have to do any see track the phase I wish the maximum let me stress agar so it's a real data safety place so what you might not do for the margin of safety is to invert the color bar
2:42:35
they're lower low hard to come read right so I can show revert spectrum
2:42:47
kind of annoying
2:42:57
all right and then two three elements for ultimate
2:43:07
and my margin that's because I have 10 Hertz so presumably if I go here gathered 80%
2:43:23
oh I see here at expiration ask for
2:43:33
now it's not going to format so what I'll need to do is I'll use a Sears or a prose
2:43:45
here's my migration and it does aah per second squared but those are definitely easier so this is something we need to handle better next version I think that's done by because we're using it
2:44:16
Digital vectors so what I know is if I want to run this with that residual vector like I can on the sign processor and I can also just frequency you can pick on your mouse
2:44:33
right like answer let's like a dude
2:44:38
of what it would be I could just clone this solution here
2:44:46
I can edit this baseline
2:44:55
and then my ever and I remove
2:45:02
let's fight
2:45:05
want to save it white
2:45:14
and what I'm going to do is I'm just going to point the nun all right
2:45:24
and then I just the new oppa - Wow
2:45:33
I only have ten Moton
2:45:40
you
2:45:42
okay yeah so let's just go down my throat uh sir just go and grab that click open
2:45:55
right
2:45:57
but it's like a joke
2:46:05
the eye in film that I was using before
2:46:12
here and ready to run okay
2:46:23
so what's the instead of not using residual vector in 40 did you call it fine respect
2:46:42
money generated
2:46:52
you
2:46:54
and we look at the different sexes before after
2:47:03
what
2:47:06
all right
2:47:18
so here I'm old I sit for this guy here put it in geez
2:47:23
one that I haven't convert I think it's a bit of a mess down here and
2:47:32
and the low frequencies here if sorry the bottom of this plot
2:47:41
not really
2:47:48
you know 1.2
2:47:52
so for this there's not much different maybe there's some more
2:48:07
you
2:48:13
okay that's a bit of French so that's a what you know easily can print the difference between two solutions
2:48:32
and with the D map I can run this with a
2:48:37
or with the and only approach and see how that compares
2:48:48
I won't do that so that's the sign Scizor
2:48:54
and it's right now
2:48:59
no sir so what you've been using
2:49:08
you
2:49:13
processor
2:49:18
our chart of dynamic excitation so we're moving from deterministic who non-deterministic if you will a random we call random Stannary process so that doesn't change change it's the the statistical properties don't change with time and I will look at what that means in a few C so much the same name as for harmonic loading with a son processor the random processor reads an ass trance Ellucian 1 0 3 solution valuates the response to subjected basic citation this time random with
2:49:57
111 except much faster and much more accurate ok well one of the nice features about it though is that it does give the same and same for the sign processor if you look at the component of the same component as an asteroid just because we're using the nester and eigenvectors right after we go
2:50:26
different results or you will for fun meeseeks because you know we viewings differently but for the conus of soil
2:50:38
all right so that's good and murder act speaking of comparison here's one I so we're looking at Sigma xx for a particular element and we rang the National result with the SI toolkit resulted
2:50:55
you
2:50:59
so on the Terminix so I mean you know you don't really know what the lane to to the responses at anytime but sufficiently by looking at the mean the standard deviation is a cup we can work with that we can get meaningful result the a is a zero me and Josh in canary or got it where the process so Tina and I guess that's why I've seen this question before the pond in the used period that steam is can I preload or can I put adding offset for the random for random answer there
2:51:49
is no I mean it assumes all the theory assumes zero mean if you offset the mean then the basic theory that's used to evaluate the response which is no good so I'm not sure those are but I I say wouldn't do that things that that the statistical properties are in a bit of time the only means that if you pick any sample la represents well what the ensemble of sample does
2:52:20
questions the pollution that doesn't change as function of operations so if excitation is Gaussian so will be the response advances that we have are good and they look like the property density function for a zero mean Gaussian to prove the mean from the equation is related to the standard deviation Sigma using this way
2:53:03
okay look here if we've expressed the the x-axis as a multiple of stations or this buddy here which is the ratio between P and the standard deviation or the RMS so here's my deviation plus one standard deviation plus two standard deviations minus two
2:53:26
so the problem of the VEX lying between minus 600 standard deviation and plus these times the standard deviation it's finding the probability this Z function from Sigma 2 plus 2 Sigma
2:53:47
so stealing energy here so the problem of being here
2:53:56
we remember top my head I don't recite right after Tucson about 60%
2:54:05
the confidence of the chance of being a big response that's worse than that and twice the standard deviation for three seconds it's 99.73%
2:54:20
and then he like
2:54:28
you
2:54:32
that's evaluated love to remember so that the transfer function we looked at before I'm for sign actually it's this is function will be responsible entity the input power spectral density times the spectrum which is now above multiple of the eigenvector is their mobile participation factor the transfer function of your transfer function their ice rink apart what it is versus the integral of this of this responder spectral density from the or limit
2:55:13
to the upper limit and those you specify in the EU i as you source the input the input power spectral density an answer the root mean squared well is the square root of variance
2:55:27
it sports the integral here means that you know doing this in a very poor way you'll be integrating each response none of the things that we don't do an essay toolkit we integrate on that on a reduced basis which which is why we're so quick and so we're able able quickly to get you another mess result which you'll see the peepers on as well and benching it about Gaussian distribution so one Mises stress is an
2:55:58
example of a function that doesn't follow a constitution because it's the square root of sum of the squares like Mike Schuh equations here for the 2d state of stress
2:56:15
and so tribution for that I mean you know give you correct results we have to know what it is okay big messes one you can see if you want you can use a Monte Carlo simulation for each residual each roulette it's slow I'll recommend you that but anyways you can what we do is we generate n samples when we sort them in descending order in this case one seven three percent confidence which is for a Gaussian distribution then we if we have an samples and that's the nine thousand nine
2:56:59
hundred seventy third tell you that's value and then we would plunder on the graph and at the distribution for zero the mean is expressed by this function here is not zero because the one makes you stresses are all positive here is by this region here and the RMS by this one here and then we plot three tigerman we dash line on the right and the peak of any are the actual peak and threes I RMS which is what people do in industry so you'll take care of
2:57:40
you take your eternal result or your response and result you think the one Sigma or the RMS result you multiplied by 3 multiplied by 20 percent you're off because the whole you know it's a Gaussian distribution you assumed it was um we public for the day fans will conference I last June I give the referee and I'll be glad to send it to you in which week investigator and making that assumption or antenna most similar to the one that I'm using here and
2:58:11
we're up to three percent error some case on the good news uses the approach from NASA and you're conservative you know anything but you're just you've eaten up your margin unnecessarily
2:58:28
and our metric which is not won't give you same distribution which is the tie rule failure in that some you guys use composite
2:58:47
I
2:58:48
beer case but Maybin day okay are quick to do it for a for random because then in processor will give you peak ply stresses animal metrics so failure indices strength ratios margins of safety to our knowledge no other software does that it complicated because if you look at the failure in this equation for so you've got some quadratic terms you got some order to to order one terms as as well in terms of the stress tensor
2:59:29
and then the stripping more involved because you got to solve for the roots of this second-order Asian by substituting the stress component at a while hey so I'm trying to solve for that so when we're handling that accurately I use Monte Carlo simulation like this one me see stress to establish the distribution and we know that our approximation is a derivative approximation this is what we do up for von Mises stresses the
3:00:12
main reference is the single Minh paper I mean this is the this is in there you probably have a copy of it and Monte Carlo on every life model because it would not be performance enough the approach of the foam approaching industry is the signal method which is the Mona Lisa stress is transformed into at most five independent central senses and they're ethically distributed Gaussian processes unit variance and they're an improvement studied in depth we know how to compute values
3:00:54
the probability of the vibration stress exceeding a certain value is defined by this tool here which is a volumetric integral up to five demands it will probably solve it it's climbing to valuate this American does invited more of a fast approximation methods that as the word this is life approximates the integral and his master step ahead of the single one method and if you want two methods a Monte Carlo snowman in fact approximation the first is the stick over to or
3:01:43
analytical the Prasanthi Carlo is excellent when you have lots of samples signals good attention to overestimate the von mises stress whereas the fast approximation I've got good Pretender estimated a bit monthly course very slow when you have lots of samples and it's accurate somewhat as fast and as fast approximation is ten times faster than signal in terms of parent yet we offer parallelization
3:02:14
for the signal across approximation methods invitations they're the fast approximation only works for high confidence levels so if you're ninety percent you should be using signalmen typically entered is three sigma 9.73 percent so that is you're okay to use the approximation
3:02:41
so what other does our AutoCAD allows you to do three load cases in a single run residual flex a residual vectors in the same way as a sign processor does I won't go to that again you can pick them use in the model superposition it computes of displacements accelerations velocities optionally it's guys hybrid integration method and the NS version it's paralyzed so that it's very precise and because of the hybrid integration scheme does not you don't have to
3:03:23
specify additional spectral lines or worry about having too late or too few I'll handle by the integration scheme so you can better at night I know what Joey and well we output the power spectral density functions the peak values so the failures are up to the confidence you specify they essentially are like three low values for Gaussian distribution you don't have to multiply your RMS is by anything we give you the peaks we're positive 0 crossing go into that a bit later
3:04:04
the prepuce exact peak von Mises stresses and margins and see from those macy's dresses for the elements that you select name methods that we just described for composites it computes Peaks failure metrics so failure strength ratio margins of safety for 2d and 3d elements so using either the P comes become GOP combat physical profit a bonus we output processor out which results in op2 and be reversal file format so in full post-processing are using an X and the question
3:04:46
was asked by Dave will the next say five work and shop verified it would again this is using advice and so it's just a real way of okay I said don't be too format guarantees that edit can can process the result the read binary Universal files as well as for the margins of safety data set because I'll be to file does not have a data set for margins and see
3:05:18
number five less you essentially it's like an apparent frequency so equation essentially I'm going to the equipment but you integrate Omega squared times the power spectral density as opposed to just the power spectral density and you take the Rideau again is your frequency for each result and that people use in durability now too so because the actual current of each result
3:05:50
staying here I guess a lot of times in the past we've talked you look at the modal responses and say oh here's the major residence and you use it for the different positive crossings but this is probably more accurate because it includes a overall effect of all the residences there right right exactly exactly if you've got I mean you've had a single mode that's creating all your stresses or your displacements
3:06:13
okay with the method you describe but you now have more than one mode that difficut energy then um you know put accurate so mine is this this result are just freely available
3:06:30
so they look a lot for the mm for the model effective mass and mass approaches exactly as a matter of fact like the one for the sign processor except that if you get asked for distance acceleration those velocities in binary format you need prime mu o ug chord countable and the rest is exactly to say so I'll let the user interface it's split up rather yeah I think in terms of the your work flow so you find your base PhD spectral density pick the freeze that
3:07:08
you want to excite definable damping so now also parameters such as confidence level that's much in the same way is in the sign processor you defined your stretch cases right so these cases you will get the von Mises stresses and safety if you just ask for stress output in the at request which is the next box you will not get the von Mises stresses you need to create stresses the base toleration dialogue will gonna create two sub cases and you can read in it you can define
3:07:46
your curve using a set of points frequency and power spectral density you can sale the power spectral density such that you can express it in in G as a function of these which is which is typically this tree here's that you don't specify output requests on this dialog all the output requests you define are applicable to all the sub cases and they're defined in a subsequent dialogue in Breezies you can follow your modes are a subset of modes whether by frequency or
3:08:22
mode ID or your mole salute position human residual flexibility effects use checking on this toggle when you're using the DMACC approach absolute and relative we talked about that before and this is what you want to define your upper at the bottom of the dialog here your lower and upper bounds for integration you can function from 22 mm Hertz say okay only integrated say between hundred and two hundred Hertz you don't need a fire on your function your modal damping
3:09:00
is defined exactly as fine processor there's no difference and is expressed as an amplification factor Q you can run all three axis at one time but if you have different modal damping for the three axes and you really have to run them individually X than Y than C that's correct
3:09:25
they don't define your confidence level so since we calculate the P values the results are non Gaussian we don't say we don't I like to do 2 or 3 Sigma you buy the confidence the confidence you can type in as a percentage or you can turn a peak to rms ratio and we will calculate Dow for this level on that and that's what we're going to use in our evaluation of results and all comfortable again is that the growth region standard deviation is letting Michael
3:10:05
so hey one thing my or RMS results you can if you warrant X but you don't need anymore
3:10:16
raishin scheme mission before we have advanced integrated integrators such that I don't need to specify the frequency content frequencies for numbers point double precision arithmetic and a false scheme of selects the most precise integration algorithm either animal or numerical analytical is exact numerical integration uses an adaptive and robust algorithm the invention is the same hybrid except there is no as it's completely numerical not there's no identical for
3:11:00
of the foot stress cases
3:11:08
this day is the flash approximation we counted approximate wannabes these dresses and nation is that little plus sign there the bottom equation there it's worth and we've done worse case Bonamici stresses and we'd give you the option to tell me accurate Vaughn may see stresses when that's high you don't know that anymore but it's there for legacy reasons if use this approximation you don't really need to worry about that at all
3:11:45
so what is your face
3:11:50
okay
3:11:56
the signal is fast or fast approximation on the bail out there again say only if you have low probabilities like one Sigma Monte Carlo you can use that you have to enter that toggle in the definition file by hand this case dialogue business sign processor you select the women since you want to process you define a length and safety factor the bus solution if you're using fast approximate approximation use
3:12:29
exact only if you're using a Dinkelman you can use worst calculates the worst case martartu lap lisa stress and then when the stress is below this margin of safety cutoff will turn on the exact time the signal message
3:12:51
less stresses the margin the custom margins below this cutoff here will be printed they want all all the stresses sort of safe out here
3:13:05
a multi request that works on cases you've defined sister must be available Naoki to file a design processor the corner requests are supported window before and now that you can look at the centroid or normal result so we can get your choice
3:13:30
to send the corresponding martyr the safety to are sent to at file excel file and hold a binary
3:13:39
per second this place and velocities again going in the displacement result in the sin they saw one if you wait to the last five and ylim ask question yeah a corner now versus corner yes the difference could you
3:14:02
is that such a real it's not a reference time processor corn reason laws of the noble results stresses I have to say yes it was also also get the sensible results in other words of those say yes I also get my elemental centroid all right and it looks at the worst of all models so he's one result for element which looks at and that result is the worst case result for all but for the centroid and the quarter
3:14:38
which are not oh yeah
3:14:47
knocking for scale factor here for the acceleration results so you can have them in G's that does not apply much it does not I into the binary results of go join X those are always inconsistent
3:15:04
um I just mentioned then you there's a composite failure metric with a selector for failure to see strike ratios and margins of safety again the stress eigen vectors must be yoga to file train as your failure theory the strain eigenvectors have to be there
3:15:26
the strings are output using the stress strain out the press directly in don't be too far up there in the club cards or in the another card for that shock will know but I
3:15:42
not that said but anyway they're in the every five interns auditing the plotting options you have don't affect at all the results that effect I look and then I select the natural [ __ ] additional frequencies using the correct one for two for three Aaron knodel results or corner results before they were ribeye and toolkit
3:16:23
the oldest day war the corners were evaluated but then they were written as an elemental results I know through have elemental no results whereas the result or only evaluated and they do not mix them the rest so the central before and after v6 are different much better
3:16:58
atom processor we have put an HTML page with a directory listing your group to excel as well
3:17:08
other side we ran that when you that you Valley open your some file and make sure that your input our spectral density is correct we should assign as well not your your input your just genetic evil wizard responds and makes exactly the same as what you input something use it errors you'll notice in there
3:17:42
um the fineness of contours of accelerations contours of stresses
3:17:51
when you see actually appearing in EDX that's not true it's our peak results it's just that we're using the root-mean-square data blocks we don't have actual voice they are
3:18:03
confidence you define and read the read the functions the departure that actually causes functions by put time and cetera to a universal file you point to the op2 for the results are in the opt file directly
3:18:22
for a demo so ideas I'm gonna read the solution it
3:18:35
you saw one every solution
3:18:47
okay stresses displacement you remember that plot dress off but my boss only
3:19:05
recorder here
3:19:11
and then we don't really need oh yeah we in deep our immunity cord level right because I'm going to ask for
3:19:26
I'm going to ask for acceleration
3:19:31
Oh
3:19:36
you
3:19:40
okay
3:19:52
sure there
3:19:58
I just got a guide my constraint
3:20:07
when you should solve it
3:20:15
basic InDesign processor except for the extra power this case specifying any residual vector
3:20:26
I could if I wanted to
3:20:38
going to open this
3:20:45
like why failure indices my stresses
3:20:49
mode shapes PC for
3:20:53
so let me use a random process sir finding a power spectral density and if I am is going to read a text file my directories is randomness and
3:21:18
you want it looks like
3:21:25
you
3:21:37
well it is a limited to columns frequency and PSD and because IP has E and G squared for Hertz while I got to put a scale factor right here and my model then I can square check for three hundred eighty six point one and is able to introduce my hair so I know three thousand Hertz bingo I'm always saying
3:22:06
curve looks like
3:22:12
but another one to say another subcase excite me
3:22:20
important
3:22:30
and just copy from one to the other
3:22:40
in fact my most except I'm just going to go from left into 100
3:22:51
damping to 25 I'm getting as a hybrid by default my confidence is going to be 99.73% 3 Sigma
3:23:06
I'm going to information to get my dresses another created my feed
3:23:18
this one
3:23:24
you
3:23:31
okay no no margins of safety of course I'm going to the binary file
3:23:43
let me refine failure metrics I'm going to put into the very fire
3:23:51
I'm not going to use groan unfortu the flyovers my safety faculty factors to do this for the selector only
3:24:05
you
3:24:07
animation of
3:24:12
okay so let's put it in the binary file
3:24:21
what for everything you
3:24:34
rexella for my response now
3:24:44
Oh
3:24:50
yep done
3:24:56
so what can he well first look at some of the first data
3:25:02
I've got it on
3:25:06
find professor
3:25:13
let's I pick the size wrong
3:25:19
you
3:25:34
I've got results cases you can see that in the navigator
3:25:52
and look at them there I've done that before so I won't bring the bleedin do it
3:26:01
noodles is just go see nice try don't be too sore to results here
3:26:09
that's actually
3:26:15
the result of
3:26:21
accelerating inch per second squared squared for Hertz
3:26:30
let's change that if you just edit it
3:26:39
you
3:26:41
blames active custom G so we put any of you to files or wasn't consistent use and you can modify
3:26:54
ah you keep off of this rel tear in an ex-gay melted use their the RMS
3:27:10
right here
3:27:14
so they're very low aren't they that did I believe you put the acceleration of gravity and your and your kickoff request did you gravity mean geez here sir watch your inline desert your inline their base acceleration is in the x-axis so you do three dishes yes that's four G's are necessary this is okay but that's bit less and this is that virtually so here our best value is here and you have them in excel file to get the
3:28:01
money into EDX if you go down if you click information on the plot you won't see it in so you just go effort if you export these is an FYI games anyway if you call Honeywell
3:28:26
take anything
3:28:35
you that's the folder sitting here I pick you here for
3:28:49
to the wrong
3:28:55
that's what an ASU for some reason I can click it go information and you'll see the upper tier
3:29:13
okay so you can't applause in that act it's just that you have to convert the plot you have to take them from the op2 file and generate records of them
3:29:29
alright nuts bro - results in that access is easier than my companion said as a matter of fact while I was like I just can use solution
3:29:46
leave it at one
3:29:50
and this go 60s go special by and I'm just going to file is generating
3:30:06
you
3:30:11
is
3:30:16
davonne makes dresses here
3:30:24
really
3:30:30
my mom a sister
3:30:34
for each ply
3:30:46
that is why
3:30:51
this is really amazing that you can get this metric solution unfairly indices this d.edge allowed interest rate ratio the ply strength ratio which is how much plus a preserver factor so you want this to be bigger than one
3:31:11
and then sodium s acceleration
3:31:18
in x1
3:31:22
and process as well
3:31:28
right
3:31:30
hey varied between cloud this is more like five cycle goes all the way to 61
3:31:41
look at you more
3:31:53
that was right now you've run big models and you've generated these statistics so it really find lots of quick wise random processor so quick we have standing up there so I got an efficient op2 reader it's got a vast approximation for von Mises dresses apparel buy it uses very far point integration scheme put on at a point you know you get something that's really exceptionally fast
3:32:32
I'm accurate it's accurate because you can account for more truncation you can have to work having enough or too many Quincy's be clear metrics and your peak von Mises stresses are accurately coded you know they're not Gaussian distributions and they've been valid using the Monte Carlo method
3:32:56
so I want to show you today as I make that come with the with the product as part of the house package and I will do them well we will redo them and we will send them to you so you can practice with them but in the night I hope you can crack if you have any support questions send them to support at my HTT calm preferably email or me
3:33:29
shaaka yeah I feature the v7
3:33:39
and I can tell you the we're not supposed to talk about the future but I think the sign processor is going to be going to be come to support not for example results it's going to be a better product pile and cranes that that we're looking at so um actually when X open script it does it we not at liberty to distribute it right now unfortunately okay so hopefully hopefully v7 this is this is smart just I won by the way of getting them set
3:34:22
Oh
3:34:27
you
3:34:34
again Wynn case thanks yeah I talked to ragel and so everybody can start seeing loaded with it for us it would to the right as Mike said as a matter of a day for us to generate the licenses so yeah my passions to accommodate are the are to accommodating everybody tweeting to temporarily I think all right we take us a day to turn that around
3:35:13
I guess wanted you guys to make a rest for Dad oh this fire email telling unwanted and we'll turn it on so many we need those are limitation on how many you give us tip all right when will they last huh also
3:35:35
that's good here how long do you guys think you need to either I think you're just good I don't know you got pocket I would like yo in it and had anything I need to do to make a bust or whatever we need to do so with this better come up with I don't know a couple weeks something like that and and hopefully it's okay for you and then we'll give you the go-ahead we'll turn that around thank you up really bit so I think we're quits are we good
3:36:29
yeah guys I just have a question I noticed you know earlier on you mentioned that the vaccination can sometimes be optimistic to get ready any techniques we're seeing that
3:36:45
hi it's that's about 2 or 3 percent how many of these right hey red-eyes with this dress yeah so what if you noticed there in certain certain cases like if certain types of models tend to make it more on a sticker you know you can like that okay talk about that I definitely okay with you so you run the analysis yeah your stresses and you apply your factor of safety and oh you're within 2% you better do it slowly so well with fast approximation you don't
3:37:31
need that so if you toggle on exact it's not even going to calculate the worst-case funny sees it's just going to go use the fast approximation for all your von mises dress optimistic potentially render you and go to the other method for the tip or with interesting hold I read it oh that's the point so I hadn't caught that and then I yes yeah you could consider going to a signalmen for myself I can group are conjoined or you just tuck on two percent
3:38:14
so it might be that you're dead on redesign then before you go redesign you'd probably do these ones yeah if you've noticed anything like you know say like larger models with lots of metallic structures tend to have our beading than say a composite that start may be more optimistic or something like that able to deduce of that high level of it's more it's more what your stress tenser looks like so I yeah see
3:38:53
what John is able to going to email everybody this presentation or you want to email me to distribute everybody really you okay just say and I'll I'll body them to sign up and we'll put it on the shared drive also and you also have recording if you would email it to me and we'll deal with it
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat