Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Largest increment in Testrig 1

Status
Not open for further replies.

BlazPlestenjak

Mechanical
Dec 1, 2021
11
0
0
SI
I am working on a model in MSC Adams Car. I have designed front and rear suspension for formula student car by myself and used FSAE templates for steering, chassis and wheels. I am trying to run Full Vehicle Analysis, but I get the same error everytime: The equation with the largest error was model.testrig.sse_diffl; The variable receiving the largest increment was model.tistrug.sse_diffl (shown on the picture below. I have tried many simulations, from acceleration and braking to cornering and open loop steering events. I always get the same error. I should mention that I currently dont have powertrain in my model if this affects the issue?

error1_aas87h.png


The only simulation that works is Kinematics and Complience Test. In this test I always get an error about ac_spring and ac_damper missing, but simulation works and I get results.

error2_dl9msb.png


I would like to ask for help as I am pretty much stuck.
 
Replies continue below

Recommended for you

So get a fresh copy of the standard FSAE model, and check that that runs whatever tests you need. Then exchange subs one at a time to identify which ones are broken.

"The equation with the largest error was model.testrig.sse_diffl; The variable receiving the largest increment was model.tistrug.sse_diffl" These error messages are not actually very useful, typically, as you tend to get a cascading failure to connect parts together that ends up breaking the model somewhere, it just so happens its the testrig this time.

"ac_spring and ac_damper missing" my models don't appear to have those components, perhaps you could look through the adm of the standard fsae run to find them.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Incidentally, a couple of things. Firstly the learning curve in ADAMS steep, and ADAMS is probably overkill for FSAE. If it is something you want to pick up for life after uni then go for it.

Secondly ADAMS models are not reliable unless you have physical test data at both a component level and vehicle level.

Third the FSAE sub on reddit has input from more than two ADAMS users.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Status
Not open for further replies.
Back
Top