Problem: Solid Bolt Pre-Load with Symmetry

Phenom
Phenom

Hello experts,

 

I did define a solid model with double symmetry and Solid bolt pre-load (preload) in NX10.

After calculation I must realize that the new artificial (decoupled) grid points defined by means of bolt pre-load violate the symmetry constraints.

 

2019-06-04_094616.png

 

 

2019-06-04_100846.png

 

Is this a bug or is this a feature?

What shall I do to have the symmetry constraints at the artificial new grid points too?

Is there a solution in higher NX Version?

 

Who can help? Any suggestions are welcome. Best wishes, Michael

| Production: NX10; Development: VB, TCL/TK, FORTRAN; Testing: NX12 | engelke engineering art GmbH, Germany | 2019-07-08
| Kudos for good posts! And if my post answers your question, please mark it as an "Accepted Solution". Kick Off (2015-09-14)On the Map (2016-10-21)1st Gear 2018 (2018-02-19)5th Birthday! (2018-02-25)Popular (100, ‎2018-10-30)Prolific (100, 2019-02-27)Philosopher (500, ‎2019-02-28)Problem Solver (50, 2019-06-25)

9 REPLIES 9

Re: Problem: Solid Bolt Pre-Load with Symmetry

Gears Phenom Gears Phenom
Gears Phenom

Look like solver slice bolt elements on two parts and create duplicate nodes to apply preload, and these new nodes don’t “remember” constraints of parent nodes.

Re: Problem: Solid Bolt Pre-Load with Symmetry

Gears Phenom Gears Phenom
Gears Phenom

Read in User`s Guide about bolt preload and didn’t find any clues except this, at the end of bolt chapter.

Снимок.JPG

However, I cannot figure out is it helpful or not. Anyway, these options are hardcoded and I think that nodes behavior is hardcoded too.

Re: Problem: Solid Bolt Pre-Load with Symmetry

Phenom
Phenom

Hi,


@Karachun wrote:

Look like solver slice bolt elements on two parts and create duplicate nodes to apply preload, and these new nodes don’t “remember” constraints of parent nodes.


yes, that's exactly what happens. So, now I ask if it's a bug? Shall I have to make an IR?

Best wishes, Michael

| Production: NX10; Development: VB, TCL/TK, FORTRAN; Testing: NX12 | engelke engineering art GmbH, Germany | 2019-07-08
| Kudos for good posts! And if my post answers your question, please mark it as an "Accepted Solution". Kick Off (2015-09-14)On the Map (2016-10-21)1st Gear 2018 (2018-02-19)5th Birthday! (2018-02-25)Popular (100, ‎2018-10-30)Prolific (100, 2019-02-27)Philosopher (500, ‎2019-02-28)Problem Solver (50, 2019-06-25)

Re: Problem: Solid Bolt Pre-Load with Symmetry

Phenom
Phenom

Hi,


@Karachun wrote:

Read in User`s Guide about bolt preload and didn’t find any clues except this, at the end of bolt chapter.

Снимок.JPG

However, I cannot figure out is it helpful or not. Anyway, these options are hardcoded and I think that nodes behavior is hardcoded too.


thank you for your contribution. The method of bolt pre-load I already knew, buit reading it once again can mostly help (also for others). The question is, if is it on purpose that the new grid points don't maintain the constraints?

So, is it a bug? - I believe it. But maybe bolt pre-load is not compatible with constraints?

 

Who can help? - Best wishes, Michael

| Production: NX10; Development: VB, TCL/TK, FORTRAN; Testing: NX12 | engelke engineering art GmbH, Germany | 2019-07-08
| Kudos for good posts! And if my post answers your question, please mark it as an "Accepted Solution". Kick Off (2015-09-14)On the Map (2016-10-21)1st Gear 2018 (2018-02-19)5th Birthday! (2018-02-25)Popular (100, ‎2018-10-30)Prolific (100, 2019-02-27)Philosopher (500, ‎2019-02-28)Problem Solver (50, 2019-06-25)

Re: Problem: Solid Bolt Pre-Load with Symmetry

Pioneer
Pioneer

Hi,

 

does the following help?

 

Our NX_NASTRAN development team has investigated PR number 9078210 and determined that the problem you have reported has an acceptable work around as described below.

 

The problem where the bolt cut is incorrect has been investigated and the following has been found.

 

In QRG, under "BOLT" bulk data entry, remark 4 states that grid points defined in the bolt should not be part of any SPCs. The reason for not supporting SPCs on bolt grids is stated below.

 

Prior to NXN12.0, the cut section method of preloading bolts (BOLT with type 2) used an MPC approach where the cut section grids were MPC-ed with duplicate grids that are created. When these grids are also part of an SPC set, the solution will not be able to solve. In 12.0, MPCs were replaced with glue at the cut but the SPC restriction has not been lifted since the same grids that  are SPC-ed are also glued. With the MPC method, the module GP3 would write a FATAL message saying the grids are part of both the M-set and S-set which is not allowed. With the glue method, there is no FATAL message but the solution may not produce correct results, as seen in this PR because of the SPC and glue combination. This will be addressed in an upcoming release but as it stands, SPCs are still not allowed on the bolt grids at the cut.

 

Workaround possibilities

  1. Remove the symmetric constraint on the grids that are chosen to be part of the bolt plane. This will get rid of the problem stated above. This would, of course, mean that the sliver of grids is not symmetrically constrained.
  2. Change bolt from type 2 to type 3 (force, displacement or strain on 3D elements) which is an initial strain based approach. In this approach, select a few element layers around the middle of the bolt shaft and apply the preload. Both bolts need to be type 3. This approach should be able to solve the model as there are no restrictions regarding SPC.

 

best regards,

Cyril

Re: Problem: Solid Bolt Pre-Load with Symmetry

Pioneer
Pioneer
P.S. I opened this call for a SOL 401 in NX12.

Note that the sliced bolt with preload has no problem with SAMCEF solver, according to GTAC support team.
Cyril

Re: Problem: Solid Bolt Pre-Load with Symmetry

Phenom
Phenom

Cyril, thank you for your answer. It's funny. My model is solvable. I can see the correct force summation for grid point element nodal at cutting section and have no fatal errors.

Maybe because of AutoMPC=YES? I'm not sure if the difference between SOL 101 and SOL401? Additionaly, I work in NX10.

But, I didn't notice the part of documentation about SPC in BOLT as yet.  - Is it working by accident?

 

Best wishes, Michael

| Production: NX10; Development: VB, TCL/TK, FORTRAN; Testing: NX12 | engelke engineering art GmbH, Germany | 2019-07-08
| Kudos for good posts! And if my post answers your question, please mark it as an "Accepted Solution". Kick Off (2015-09-14)On the Map (2016-10-21)1st Gear 2018 (2018-02-19)5th Birthday! (2018-02-25)Popular (100, ‎2018-10-30)Prolific (100, 2019-02-27)Philosopher (500, ‎2019-02-28)Problem Solver (50, 2019-06-25)

Re: Problem: Solid Bolt Pre-Load with Symmetry

Pioneer
Pioneer
Not sure there is something about slicing bolts in the documentation. Glad anyways that you could solve your problem. Smiley Happy
Cyril

Re: Problem: Solid Bolt Pre-Load with Symmetry

Phenom
Phenom

Cyril,

Oh no. The artificial grid points have no BCs. So the problem is still there.

 

But, the simulation does not abort with FATAL error because of the definition of BCs at grid points belonging to the bolt pre-load definition too. Now I'm not sure if it is a good way to define bolt pre-load in symmetry areas as NX forbids it in the documentation of BOLT

 

 

2019-07-02_095912.png

 

So, what is the better way?  - Accepting that NX solves the solution with tiny deformations in symmetry definition and hoping the best, but, having a model conflicting with the guidelines with the chance of other inner errors with unclear inner behavior? - OR, don't using bolt pre-load in symmetry?

 

I don't know. - Best wises, Michael

| Production: NX10; Development: VB, TCL/TK, FORTRAN; Testing: NX12 | engelke engineering art GmbH, Germany | 2019-07-08
| Kudos for good posts! And if my post answers your question, please mark it as an "Accepted Solution". Kick Off (2015-09-14)On the Map (2016-10-21)1st Gear 2018 (2018-02-19)5th Birthday! (2018-02-25)Popular (100, ‎2018-10-30)Prolific (100, 2019-02-27)Philosopher (500, ‎2019-02-28)Problem Solver (50, 2019-06-25)