cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Showing results for 
Search instead for 
Did you mean: 

Community Tip - New to the community? Learn how to post a question and get help from PTC and industry experts! X

Help with modeling hammer impact to simulate pyro-shock

cmitchell
1-Newbie

Help with modeling hammer impact to simulate pyro-shock

Our senior capstone project involves building a test rig with a drop hammer to exite a fixture that will mount electronic components to simulate a pyro shock to them. We have creo avaliable at our school, yet we have no experience with simulate only with the CAD part of creo. To begin we don't really understand where we should start to be able to get some simulated results to improve our experimental model. I have a test plate fixture and we can do a static analysis using a pressure force to simulate the hammer impact, but I know this is going to be vastly different given a dynamic impact. I can also run a modal analysis, but only with a low number of modes, if I try to select a range of frequency simulate runs forever. I can also do the dynamic analysis and do the shock part, but don't really understand what its giving me.

So my question is, does simulate have the capabilities I need to be able to vary input parameters? (impact force, impact location, boundary constaints on fixture plate) In order to see what effect this will have to improve the experimental model. Im also not sure exactly what outputs simulate will give us? (an acceleration response of the fixture?) It would be nice if simulate would actually give you a Shock reponse spectrum given a certain impact? Our project goal is to be able to correlate the test inputs to the acceleration output and corresponding shock response spectrum. They would like a more systematic and efficient approach in lab testing to be able to know what input variables they have to put in to get the correct shock output they are looking for.

Thanks for the help.


This thread is inactive and closed by the PTC Community Management Team. If you would like to provide a reply and re-open this thread, please notify the moderator and reference the thread. You may also use "Start a topic" button to ask a new question. Please be sure to include what version of the PTC product you are using so another community member knowledgeable about your version may be able to assist.
11 REPLIES 11

Ive been able to make some progress I think. Ive got an mechanism analysis made with a playback where I drop my hammerhead onto my fixture plate using gravity, and 3D contact surfaces on the hammerhead. I go to use in simulation in mechanism, and select the fixture for the body and components and click ok to export it. I save it but when I go to simulate and try to load the load it says theres no loads to import for this part/assymbally?

Hi, Mitchell,

Could you attach images or file your hammer impact system.

Best Regards.

Denis.

Here is a link, I don't see where I can upload a file. Its has 3 files which im assuming can be opened for each version of creo. The verison I have isnt compatiable with my school computers.

https://drive.google.com/folderview?id=0BxZLxR3BqU2FYWp2U01MTUNpWlU&usp=sharing

Let me know if that dosent work and I can get the file directly off the school computer. I think the shcools are Creo 2.0.

Anyway I dont have the full test setup we are going to use since we made it in a newer version of Creo and can't open it(or at least havent found a way to). I have the fixture plate as a part and just made a 20lbm hammerhead with a cylinder joint and 3d contact surface to fall 45 inches with gravity onto the plate.

Whether you can do this in Simulate depends on several factors:

  1. Physics associated with the system.
  2. How the system is designed (assembly methods, materials, etc.)
  3. Type of results you're after.
  4. Accuracy of the results.

The more information you can provide, the better. Since this is a school project, do you have a project proposal written up or part of your final report that documents the specifics of the system?

I can also run a modal analysis, but only with a low number of modes, if I try to select a range of frequency simulate runs forever.

A modal analysis determines the eigenvectors and eigenvalues of the system; i.e. the nature frequencies of the system and the mode shape associated with each. These characteristics are critical for any type of dynamic analysis (as Simulate only does linear dynamics). A key part in ensuring accurate dynamic analysis results is preventing modal truncation, which happens when you don't capture enough modes in your modal analysis and therefore truncate off the dynamic response from these missing modes. How you determine the number of modes you need depends on the type of dynamic analysis, but a good starting point is for your highest mode to be ~4x greater than the characteristics frequency of your input load. You can also run a Dynamic Frequency analysis and look at a plot of maximum displacement as a function of frequency to ensure that the response of the system at the higher modes is significantly lower than the lower modes.

I can also do the dynamic analysis and do the shock part, but don't really understand what its giving me.

There are four types of dynamic analyses in Simulate:

  1. Dynamic Time: This is a time-domain analysis where your loads vary as a function of time. You can simulate impulse loads (like an impact) with this, among other types of time based events.
  2. Dynamic Frequency: This is a frequency-domain analysis where your loads or base excitation vary as a function of frequency. A sine-sweep analysis is a good example for this.
  3. Dynamic Shock: This is a frequency-domain analysis where your base excitation is defined by a Response Spectrum Curve, which defines how the structure responses at a given frequency. This method is typically used to simulate seismic responses.
  4. Dynamic Random: This is a frequency-based analysis that uses a Power Spectrum Density as an input for loads or base excitation. This type of analysis is statistical in nature (your outputs are either PSDs or RMS values) and is usually used for inputs that are very random (think of the forcing function associated with driving over very rough terrain).

Which analysis type you use depends on the nature of your input load and the type of results you want. As I said earlier, Simulate only does linear dynamics, so all non-linear effects will be excluded. Example of non-linear effects are:

  1. Contact between components (i.e. path-dependent solutions).
  2. Material non-linearities (plasticity and hyperelasticity).
  3. Large-displacements and strains.
  4. Geometric non-linearities (e.g. loads that stay normal toa surface as it deforms).

So my question is, does simulate have the capabilities I need to be able to vary input parameters? (impact force, impact location, boundary constaints on fixture plate) In order to see what effect this will have to improve the experimental model.

Simulate does allow you to run sensitivity studies were you vary one or more parameters (e.g. geometry dimensions, load magnitude and direction, etc.) to generate a plot for certain results. However, these can be very computationally expensive to do, so it's important that you do as much work on your end with engineering judgement as opposed to just throwing it at the computer.

Im also not sure exactly what outputs simulate will give us? (an acceleration response of the fixture?)

It depends on the type of analysis, but generally speaking a dynamic analysis can give you displacement/velocity/acceleration/strain/stress/reaction forces and moments, as well as many other values as a function of your input or the total response to your input. Keep in mind that FEA is best when you know a specific result your after, as opposed to throwing a model at it and seeing what it spits out.

Our project goal is to be able to correlate the test inputs to the acceleration output and corresponding shock response spectrum.

Depending on the specifics of the systems, this should be very doable since all you're after is acceleration (as opposed to stresses).

I'll take a look at the CAD you posted. Additional information on exactly what you want to do would be very helpful.

EDIT: It looks like your Google Drive link only contains the assembly file; we'll need all the associated part files to look at your model. What you can do is create a fold in your working direction and then do a backup to that folder (File-> Save As -> Save a Backup). You can zip that folder and then upload it to Google Drive.

Thank you, this is really helpful. We have the resources but its hard to get the professors time to learn to use them. We use creo at school exclusively but we were only taught in making models and drawings, not simulation, so it would be nice to do as much as we can in creo. I would also learn how to use credo for other things in the process.

Anyway I'm using my phone right now, but pyroshocks can be 10,000Hz or more and last 50ms or less. So you can imagine acceleration vs time (from accelerometer) is pretty useless. Our sponsor has also said computaional methods have difficulty with this high frequency type of shock(tends to overestimate). We have yet to build a physical test setup but or objective first is to hold constant(swing amplitude, impact location, boundary conditions) everything accept one parameter. I would hope to use creo to come up with an initial experimental setup and then further iterate it from there. We hope to be able to match Shock response spectrum outputs(from accelerometer measurements) using predictable variable inputs to the test setup.

I will post up our documentation later, we are still in the first half and due to the complexity of the project our sponsor has extended to next years senior class also.

Thanks again for the help, I would be completely lost without all this.

Ive been trying to import the load set from mechanism to simulate, but its still giving me the error saying there are no imported loads from this part assembly?

Here is a link where I saved a backup. I also included the playback file where I simulated the hammer falling with gravity as a cylinder link.

https://drive.google.com/folderview?id=0BxZLxR3BqU2FMVhmUmNYNHJ4TmM&usp=sharing

I can output normal force when I define a 3D contact surface by measuring the normal reaction force of the impact plate by the hammerhead on the contact joint.

While this gives me info about the force over time (about 300kN peak over about 0.001s), its probably not particulary useful in any way. Im still having trouble using the mechanism loading in simulate. I use the dialog box to select the loadings from mechanism to extract the peak loading and go to simulate and try to use the mechanism loads, but it says there are not any defined?

Im hoping I can transfer the mechanism loading the simulate and maybe find the particular mode the plate is exited at given the simulated drop in the hammer in mechanism with all the mass properties defined?

Hi, Mitchell,

As you indicate you Creo to school, it is a specific version and you can not open your files with professional versions.

Attached is a document that could interest you.

Best Regards.

Denis.

http://www.sandv.com/downloads/0906alex.pdf

Thank you Denis for the pdf. Given a proper modal anaylsis caputuring all the modal responses it seems I can get valid results with simulate.

I exported the results of the normal reaction force on my 3D contact surface to excel. I deleted the data to include only the data where their was a non-zero impact force. I then numbered each data set with 0.001 increments (as 0.001 is the time incriment I used in the mechanism analysis) I then saved this and imported the results for the time dependance in a dynamic time analysis in simulate. However I get an error measure trying to click of, it says no measures valid for the dynamic analysis have been defined for your model.

I have a load with a magnitude -1 in the Z direction since I figured it will multiply by the imported time dependent loading I have in the dynamic time analysis.

Ok so I see that I did not have my measure defined in simulate. So I defined a measure to measure the acceleration of my impact plate, but when trying to run the dynamic time analysis it still is giving me a fatal error.

Ok so I suppressed the hammer swing mechanism so the model would just contain the frame and the dynamic analysis ran. I graphed the measure of the acceleration response of the impact plate fixture and the results look somewhat valid at a peak of 31g's. Now im guessing as was said earlier that im getting some modal truncation since my peak response was only a small 31 g's and that I need to capture some higher modes since my modal analysis was only done with 6 modes.

Im going to try as Shaun suggested and look at the frequency analysis and the displacement to make sure I have enough modes, but I might try to just run another modal analysis to caputure all the modes in the 3000-10000Hz range which is the range of mid field pyroshock.

Im still confused on how to define the loads for dynamic time analysis. Right now I just have a base load applyed over a very small impact region with the magnitude of 1 I defined on the impact plate from looking at the contact region in mechanism. I also have the load I imported from excel for the impact load over time. I assume since the magnitude of the load is one it will just multiply that by the imported load from the table?

Top Tags