Reaction Forces

[Alexander]

Hello
I haven't used dynamics much and was just wondering if anyone knows if it could be used to simulate the reaction forces of an object when vibrated at different frequencies.
I am designing a range of products which will absorb vibrations and minimise resonance. So I have a floor surface that will be vibrated and I want to record the reaction forces of the products placed on the surface.

Any help would be gratefully received.

Cheers

 

[Lew_Merrick]

Alexander,

I am unsure if you are referencing the "old" Alibre Dynamics of the 3D System attempt to incorporate DST's Dynamics program. There are/were so many "issues" with the DST program that I would seriously doubt any "result" generated. I do not recall a specific vibration driver in "Alibre Dynamics" -- though it has been five years since it was available/functional. How about FEMDesigner? It has a pretty good vibration driver "action" condition. ???

 

[Alexander]

Hi Lew,
The Dynamics / Simualte which 3D Systems offered, the really clunky and difficult too use one, I recall it being real pain to use.
I forgot about FEM Designer, was this an add-on, additional purchase? I've not been a Alibre/Geomagic user for a while and stopped maintenance at V16.

Cheers Lew, good to hear from you.

 

[Lew_Merrick]

Hi Alexander,

The DST program has way too many faults to be trusted (IMO). I have used FEMDesigner for work that had to pass DoD, NASA, and the like's specifications. I would guess that a "quick" and "default analysis" with FEMDesigner is a ±5% resultant. Using more "precise" definitions gives me (typically) a ±2% resultant.

 

[Alexander]

Okay thanks for the info, I'm going to email the guys at FEM to see if they can help. It looks good and I like the Alibre integration.

Thanks Lew

 

[listastudio]

...The DST program has way too many faults to be trusted (IMO)..

The issue is NOT in the software (Working Model/visualNastran/SimWise), but in the bundle made by Geomagic. A very powerful kinetodynamical plus fea (Nastran) engine, namely SimWise 4D, offered at 1/10 of its price without any training and support, than we cannot expect CAD users to truly understand it without any help.
Specifically, if the collision model is impulse/momentum (as default), then reaction forces during collision are zero because we are solving collision directly calculating the mechanical energy wasted at any collision (such as, supposing collision closing in such a small time to be neglected).
Put a ball over a table, set gravity, set ball and table in collision, set coefficient of restitution 0.9.
Set ball and table in collision detection.
Then, with these two bodies selected, open a window to measure forces between the two bodies and another window to measure contact impulse: You will see that during the bouncing of the ball the reaction/contact forces are zero (while the impulse meter pops out the energy wasted) and only when finally the ball stops bouncing You will see the reaction forces starting to appear and then computing for the weight of the ball.
In fact, only when vibrations are ended there is time to measure forces, while during bouncing we were supposing that the time of mutual contact (to measure forces) was neglectable at every bounce.
So, to measure forces during collision another strategy should be performed (for example, changing collision model to take into account the time of mutual compenetration and deformation at every collisions).
This software (SimWise) leads to very accurate results - when explained and used in the proper way. There are several products and machines on the market developed with this tool, for example several stage equipments at the famous "Teatro alla Scala" of Milan, Italy as in http://www.lista.it/atnet/cassca.htm and other in http://www.lista.it/atnet/index.htm

 

[RCH_Projects]

without any training and support, than we cannot expect CAD users to truly understand it without any help.

Only the last 4-5 months I suppose that I have worked out the "impulse meter" as you describe it.

A cam follower on a spline can rotate a lot without a constant enough contact (at restitution 0?) to be registered in a force meter.
I hunted down a "conversion" factor that might correct for extended issues but without an opinion of it's value as I applied it, I skip it for now.

Now, if you or someone had mentioned this and perhaps a few other aspects in the beginning it might get a better rap.
Too many broken links in the provided help files right where I wanted to dig deeper (note I do NOT have/use fea).

Graphs must have Y axes scales reset after a save and reopen and occasionally other times that make it a pain, that would seem a simple fix.

I'm coming up to speed after all these years "winging it". I figured that paid support is bread and butter so a lot doesn't get thrown out there.

It is my only "lifeline" with the features at nearly no cost; except for maybe a $90,0000+ learning curve if on salary.

Figuring out "reaction torque" was an original question I submitted without knowing how to ask the question - total newbie.
If the totally useless (and rather pompous I thought) reply had an emoticon it would have been an :roll:, maybe even meant as a ..I.. off.

Now I have a separate flywheel on an axle driven by a belt constraint.
To understand the torque and prevent pulsing or runaway rotation that leads to sim fails when the steps cannot be maintained I setup high friction on the axle/flywheel and use a special RPM Limiter.


(The 775,000 number is so far an anomaly at startup)

It (appears) to be a good combo and I will eventually have "friction" turn off itself when RPM limit is running so adding the results isn't needed.
It is here I would have added the "intermittent" value if I were not past the need (mostly) for evaluating results (I may have to reevaluate that need but I am in deep already).

I eventually have a project to take up with an engineer like Lew or even Lew and it isn't good that there is any doubt about results I might present.
I don't have the engineering credentials or math background to independently counter any scoffing of a study.

If it is wrong...

Other than that I give the experience a "slow clap" but improving.

 

[RCH_Projects]

BTW!

I have been sitting here for days rerunning sims that end in a software crash (sorry you cannot have a report until R&D ends).
I have maybe a second after errors start listing out in the event viewer to "stop" the simulation and examine the states.

I am running it "low and slow" and watching :shock:, but it slips by me an hour or hours into simulation.
Refinements are helping but I am going to start the sim from initial position always after refinement.
(Yes I know how to "Erase motion history" and preserve settings at any point, when I want to "bump" a simulation, but that is only useful to me on completed models).

I can expect a few more potential "error zones" before I finish refinements and add models that will be vulnerable as well.
They do clean up.

A way to stop on event code or auto save every sim-minute or sim-second, so it can reopen within a sim-minute or second of shutdown would let me get some sleep .

Ya know us CAD users - we be a bit thick! :mrgreen:

 

[RCH_Projects]

I have care giving and rest needed - then I will come back with a list of devastating "work flow" mistakes you must not make.

(YMMV).

 

[bemfarmer]

Hi RCH,
Are you using Dynamics?
Looking forward to your informative posts on using Dynamics, as I purchased it (2+) years ago, but have yet to try it.
- Brian

 

[RCH_Projects]

I come to the forum to offer help, get help and keep up to date.

I have about 1200 .wm3 files in current folders and 5000 in backup (auto backup of changed files at 30 minute increments). You could imply that I am an experienced user.

I take into account that Dynamics is a niche product, as is Alibre. But Alibre is worked on and the users tended to (again).

Dynamics can be fussy and clumsy but functional software. Most issues are easily detected and avoided if you are aware of them. Some, where a part(s) lose physical attributes (Mass = 0) and will not reset, or the state of parts (velocity) do not reset after Erase Motion History and have to be reset in properties, can wreck havoc if you don't recognize and watch out for them. Note that if you Erase Motion History and you are not located at the initial condition, step 0, the software will alert you under normal conditions to ensure that you are intentionally setting the selected point in simulation as the new initial state.

Usually when you create a change affecting the simulation the software will reset to initial conditions. Sometimes it seems to miss something. So always reset manually to the initial simulation start, before changing a meter or adding a meter or formula changes, material changes, etc.

There are things like, you can't add a "meter" and undo. Don't delete a meter from the window - delete it from the Name list. I've had issues from these actions that I don't try to replicate or know if they cause corruption every time.

Stopping the simulation can hang in a waiting state a long time (for in-progress calculation rollback usually) and/or may cause a random shutdown by getting ahead of the software ready state by clicking in the window.

Tolerance and integration settings can determine if a model flies apart in a single step (and often becomes unrecoverable). If you change your settings - "save as" to a numbered version.

I put some steep challenges on the software. Rarely the software encounters a condition it just can't handle and shuts down. Not a Tolerance and integration issue but changes to the model have corrected it. Finding small issues like that are important before applying for IP protection. A real world evaluation will clarify the (unlikely, due to redundant features) issue but my design allows for adopting or removing the minor alteration.

All of this can be overcome with experience and good modeling. I am using tougher elements than the average user, with splines and 3D contact involved.

Professionals like Ralf, Lew and others provide what they can and have valuable insight. Time permitting I would get into the more fundamental issue of constraints and coordinates (their default locations, movements, relocating, splitting, swapping, joining and other relationships). A "fixed" constraint is the least issue.

The Alibre side of the equation is solid. I don't know how many issues are in the Alibre Dynamics version compared to the vendors other releases.

I wouldn't knock the software for these issues; but I wouldn't imply users are ignorant without hand holding by those responsible (which they do not offer). Clean up the software or give some guidelines on best practices first. If I threw you a rope and water skis and just told you to "hold on tight" that wouldn't make me a good boat salesman either. If you think this is the level/quality of software that you think should represent what you offer (because it's cheaper for us) then that is on you.

You do not have a forum or offer support so we try to do it here. We may get it wrong or cause you to look bad, but you aren't here. Get in and stay in or close your eyes and let us work it out without you!

I have a lot to do, so it is all the time I have to spend on this. I wouldn't throw the software out but don't disparage the users.

Best of all, it isn't "in the cloud".

Thank you!

 

[RCH_Projects]

Hi RCH,
Are you using Dynamics?
Looking forward to your informative posts on using Dynamics, as I purchased it (2+) years ago, but have yet to try it.
- Brian

Generally I try to keep a sock stuffed in my mouth.
Alibre is such great software and a great company again.
Sometimes that sucker blows right out though.

Just spent the night trying to activate contact between two parts. Two other parts have been running with it all along.

I've seen it plenty before.
Usually I switch through contact settings and values until something locks in then go to the settings that work best.

This is otherwise just a "revolute joint on slot" (lever fulcrum).
I've done a multitude of settings and integrator choices and watch as it slowly passes through each other.
Rebuilt parts, increase clearances, reset materials, restarts, no interferences found in Alibre, no warnings.

So Brian, keep aspirin handy.

Edit - OK, the rebuilt part geometry did not come through (saved and all) and/or only a portion of the part is recognized in the wire frame mode and misses the lever.
That'll clean right up.