Looking for opinions on the use of Autodesk Simulation mechanical (ASM) for steel plate structures 4

[barrti]

Greetings,
Could any one give me their opinions on the use of Autodesk Simulation Mechanical (ASM) for the strength and fatigue analysis of welded or bolted steel plate structures?

Our draftspersons use Inventor exclusively so ASM is an obvious choice however, we are also considering NX Nastran and Strand7.

Our key questions are:

1. Is ASM a good choice when conducting strength and fatigue analysis of large welded steel plate assemblies that have been created in Inventor as a solid model?

2. Because our assemblies are reasonably large Inventor solid models, is the geometry cleanup and meshing on a par with other similar software?

3. How have people found the Advanced Support (Phone and email)? In Australia, we are usually 12 hours out of step with USA, so natuarally we have some concerns here.

We are seeking feedback specifically because Autodesk has been unable, at this point in time, to put us in contact with Australian users of ASM who do similar work to us.

Thanks in anticipation

 

[4Pipes]

I can only give a part opinion on this.
a) Support
This was a concern when Autodesk took over at the end of 08. However, to date this concern has proved groundless - at least for me. Quite a few of the old support names are still involved. I have had replies within a couple of hours after posting questions at 8.30pm on a Saturday. This might have been an exception but it has always been pretty good.
b) Autodesk Simulation includes both mid plane meshing and layered meshing which you will need need at least in principle for your solid models. My experience with both has been mixed. I don't have experience to compare it with other applications. I have chickened out and gone back to surface models.
I fully accept that I might be unfairly criticising Autodesk Simulation. Things might have improved.
On the plus side, Autodesk Fusion is included which has surface and solid modelling. Fusion also includes a simplification tool to automatically remove small fillets and features.
I cannot think of any obvious reasons why Autodesk Simulation should not be suitable. The price should be good because Autodesk are trying to muscle in on Ansys and Abaqus. That is why I use it.
The best option is to get a working demo for a month, maybe longer. Do the same with the competition. All major players should fall over themselves to give you a free temporary licence.

Hope this helps

 

[barrti]

4Pipes,
Thankyou for this. It has been very helpful. We value the feedback.

Regards from Western Australia

 

[EngAddict]

Hi barrti,

I am located in WA and use ASM for the same sort of work.

For your questions.
1. Yes works fine for this type of work. I assume you are manually performing fatigue analysis not expecting to use the fatigue tool? The linearisation tool needs some improvements but is ok. For hot spot stress extrapolation ASM works fine. We do fatigue work to AS1210, PD5500, ASME VIII-2, ABS and DNV.
I don't like using Fusion myself but it does delete fillets nicely. There is also a de-feature option directly in Inventor or you can do it manually in a derived part using standard inventor features (e.g. delete face and heal). You also can directly edit inventor parameters in ASM but will often need to fix up a lot of your inputs. It isn't as good as SolidWorks for editing geometry and split lines with minimal to no rework.
ASM has a great surface mesher but the volume mesh creation is random and needs a lot of work. You can use a layered mesh but only for flat uniform thickness plates. There are structured mesh options but no swept, extruded or revolved options! This would make me opt for Abaqus or Femap/Nastran or Ansys (probably in that order) if I could choose a replacement software.

2. I prefer to do all of the de-featuring manually in Inventor, then just bring in the part I want to mesh. There will almost always be remodelling to some extent to get something suitable for your design loading, particularly with weldments, plates and shells, so why not use a parametric CAD package if you have a choice.

3. I still have a technical query outstanding with Autodesk and don't expect a response any time soon. I don't think Autodesk Australia have many qualified engineers on staff. You can always email the SimSquad (simsquad@autodesk.com)(

http://usa.autodesk.com/adsk/servlet/pc/index?id=16661243&siteID=123112)

or use the Autodesk Simulation forum but expect about 1 day delay. That said Solidworks and Inventor VARS in Australia usually take about the same time to respond to a request anyway.

Like 4Pipes said, expect good value for money but it doesn't really compete with the others IMP. Regarding what I said about the structured meshing, I think this will be the key to be able to produce a competitive package to start with. Being Autodesk owned though is always a good thing, expect the user interface and ease of use to increase rapidly but expect the technical features to be slow coming.

Just my opinions but hope it helps.

 

[barrti]

EngAddict,
Thanks so much for your reply. It confirms my impressions from other sources. It looks like we are going ahead with ASM in the near future. This is largely driven by the useability factor with Inventor. We mainly use Inventor and are expecting to find the interoperability with ASM to be a strong point. For simple assemblies, Inventor FEA is a quick tool to get a feel for stresses and deflections. We have found limits as the assemblies get more complex. We too are waiting on the resolution of a curly problem from Autodesk!

1. Yes I will initially be using a manual fatigue approach. I have only been looking at fatigue for welded structures this year. I've been away from mainstream Engineering for 12 years ( have been teaching instead). From what I've seen so far, I'll use a hot spot approach and not model the weld itself when I can.

2. From my limited reading of ABS and DNV, it seems like I will need to model fillet welds using solid elements and then extract normal and shear stresses to find the stress range. Is this so or can I use a hot spot approach with fillet welds as well?

3. Initially, I started with AS4100/AS3900/AS4324 however, thanks to yourself and others, Ive been looking at ABS and DNV. We dont get involved in pressure vessels. I've found ABS and DNV to be helpful. Are these sufficient or are PD5500 and ASME VIII essential?

4. Since we mainly use steel plate/RHS/PFC/I-beam,Columns and pipe sections, I am hoping the layered mesh will get me out of trouble with ASM.

5. Aside from the standards, do you have any recommended reading for the sort of analysis I am doing?

 

[ChiraL]

Dear barrti and EngAddict,

I have a question regarding Algor. I have recently started using an old algor FEA packing dating back to 1999. It works perfectly but I am unable to select the analysis type to 'nonlinear stress with nonlinear material models". How can I get about this problem?

Many thanks,

ChiraL

 

[barrti]

ChiraL,
No, I'm sorry. I have only just started using ASM and have no history with Algor. Perhaps some of the Algor users could comment or possibly contact Autodesk themselves. I believe many of the Algor crew are still there.
Cheers

 

[EngAddict]

ChiraL, you double posted here and on the Inventor forum. I replied there before I read this but I don't think you will have much luck with such an old package. Can you post any screen shots, is this the only analysis type that doesn't work?

 

[ChiraL]

EngAddict,

Thank you very much for your feedback. I have recently joined my organisation as a graduate and this was the version they had. Since the latest versions are costly our organisation is not too keep on that sort of an investment at the moment.
The version I have is Algor Superdraw III version 12.00 - WIN. To begin with I was trying out this non-linear beam tutorial, I was following the instructions and that is when I came about this problem.
The drop down menu on my screen does not allow me to select 'nonlinear stress with nonlinear material models' since it does not appear.

Further more do you happen to know how to find the shear stress of a carbon fibre tube or cylinder using FEA?

Regards,
ChiraL

 

[4Pipes]

ChiraL,
You probably only have a linear licence file. You had to pay more to include non-linear. It was only in the last couple of years since Autodesk took over that the linear option was no longer sold separately. It was a great free upgrade if you previously only had the linear version althought the subscription is more expensive. Make sure you keep a copy of the licence file.

 

[barrti]

EngAddict,
Thanks for your previous post. I starting to use ASM for a large assembly consisting of beams, plates etc. I've spent some time using fusion to delete fillets and also using the Inventor "simplify model" tool within Inventor.

I have had trouble finding an easy way to delete fillets on the frame generated beams apart from individually selecting them within fusion. The Inventor tool won't handle the fillets on the beams. Is individually deleting them in fusion the only way?

Assuming that I am meshing beams with midplane mesh in ASM, Is there a way to avoid deleting the fillets on the beams?

Thanks

 

[EngAddict]

I don't usually run a full FG model, that would be done in Space Gass, we would only do the structural connections in FEA and likely model it to suit or derive the relevant parts. However you can simplify fillets in Inventor using the delete face command and select the heal option (see attached). You might want to derive the cad files into an inventor part file then make your modifications to that part rather than the main assembly. If you want the midplane mesh to work on beams the only way I know is to remove the fillets first. I like to use Inventor for everything and avoid Fusion at all costs. You could probably simplify a lot of the model with beam elements in ASM for large models. Do you use a structural analysis program as well?

You could maybe look into Autodesk Robot Structural Analysis for compatibility with Inventor. It uses the same solvers as the Inventor frame analysis.

 

[barrti]

EngAddict,
Thanks again for this. I'm still on a steep curve with ASM!
I'll try the delete face/heal option as well as some of your other suggestions-thanks.

I finally found fusion to be reasonably ok at deleting fillets but it took me a while - it's still a bit clunky.

I'll need to look at the beam element option soon as you suggest although a lot of our models are quite short and stocky i.e. the beam sections/angles and pfc's are short so I'm hoping to use midplane meshing quite a bit.

We don't use a structural program although one of our associates does. I try to stick to the more mechanically oriented projects.