Solid Bolt Pre-load in cylindrical cutting plane needed

Phenom
Phenom

Hello experts,

 

I'm using Simcenter 10 for NX Nastran 10, solution 101.

 

Is it possible to define solid bolt pre-load at other cutting planes or elemental cutting sections as flat planes?

As in my case with a screw connection (without nut) the bolt pre-load is working at the thread of a screw.

I wounder if it would be possible to define this abstracted cylinder face between screw and surrounding structure with internal screw thread as the specific surface for pre-load?

 

 

In my picture (cutting section of cylindrical model, only bodies without elements) the screw is pink and the surrounding structure with internal screw thread is grey.

 

Momentarily, I did place the bolt preload at the starting position of shaft. But my wish is to place it at the cylindrical area between the red dots where my model is merged. The red dots mark the nodes of mesh which are duplicates and therefore not connected.

 

2019-06-11_103250.png

 

Is this possible? - Maybe in a newer version of NX??

Who can help? - Best wishes, Michael

| Production: NX12; Development: VB, TCL/TK, FORTRAN; Testing: NX1847 | engelke engineering art GmbH, Germany | 2019-08-23
| 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)
| Credo: Provide help if you want to get help!

5 REPLIES 5

Re: Solid Bolt Pre-load in cylindrical cutting plane needed

Solution Partner Valued Contributor Solution Partner Valued Contributor
Solution Partner Valued Contributor

Let me first explain the principle behind bolt preload boundary conditions, which may help.

 

The bolt preload on solid elements will effectively split the bolt in two along the split plane and the preloads are applied as equal and opposite forces on the two sides. The split plane may be at any part along the length of the shank. This is why when you are viewing results on solid element preloaded bolts in a post processor, the bolts will be split and the displacements of the two cut faces will overlap. This axial displacement in the solid element bolt does not reflect reality, but the strain and stress in the bolts is correct.

 

This method therefore completely disregards the geometry of threads, for you to best approximate them, I'd keep the split plane of the bolt shank close to the head of the bolt, like in your picture. This will keep the threaded part of the shank clear for you to apply weld-like glueing between the shank cylindrical face and end face and the hole faces the bolt is fitted in. The glue will best approximate the threaded fastening and from glue results (BGRESULTS), you can see the amount of shear the glue is subjected to. Some hand calculations will then be necessary to determine stress per length of thread.

 

I think this should do you fine, much better than the only alternative, which is explicitly modeling the threads as 3D geometry.

Re: Solid Bolt Pre-load in cylindrical cutting plane needed

Phenom
Phenom

j_coetzer,

thank you for answering. I already could see the abilities of bolt pre-load.

I did  ask myself if there could be chance to use a cylindrical cutout . Reason: Sometimes it's not possible to define a neat flat cutting section so a cone or a cylinder section would be better. The relative movement of the both sides of cutting section would also be possible by this cut out and an averaged sum force is also possible.

 

What would happen if I change the selection of GP in BOLT ENTRY for ETYPE = 2?

 

In the description there is not information about having to define a PLANE or FLAT cutting section for bolt description (except for SOL601), right? - Maybe it's only a problem of Preprocessor?

 

What do you think? - Best wishes, Michael

| Production: NX12; Development: VB, TCL/TK, FORTRAN; Testing: NX1847 | engelke engineering art GmbH, Germany | 2019-08-23
| 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)
| Credo: Provide help if you want to get help!

Re: Solid Bolt Pre-load in cylindrical cutting plane needed

Phenom
Phenom

All right, here are my results of research:

I did renumber the grid points of cylindrical face between screw and surrounding structure and did manipulate the associated grid points in BULK DATA ENTRY BOLT

 

$*  NX Load and Constraint: B-PL_1kN
BOLTFOR        61.0000+6   23243
BOLT       23243       2       0       1       1       2       3       4+
+              5    THRU     691

After that I've got the following error message in f06:

...
*** USER FATAL MESSAGE 22201 (MODGM2BC) BOLT GRID 111 IS NOT CONNECTED TO ANY ELEMENT THAT LIES ABOVE THE DEFINED CUT PLANE. THIS COULD BE DUE TO SKEWED ELEMENT SHAPES OR IMPROPER DEFINITION OF BOLT DIRECTION. *** USER FATAL MESSAGE 22201 (MODGM2BC) BOLT GRID 145 IS NOT CONNECTED TO ANY ELEMENT THAT LIES ABOVE THE DEFINED CUT PLANE. THIS COULD BE DUE TO SKEWED ELEMENT SHAPES OR IMPROPER DEFINITION OF BOLT DIRECTION. ...

So now I understand the problem. NX NASTRAN has to decide which associated elements of grid points belong to the positive or negative side of cutting section.

NX is deriving a REAL FLAT cutting plane from grid point coordinates by means of orientation of bolt force (here in X direction = 1). But my elements are placed around cylindrical face so I have different levels of grid point in X-direction. That can't work because after defining the X-position of cutting plane most of my grid points are not part of that cutting plane and therefore most of the connected elements also not.

 

But it was a short try to understand the manner. Best wishes, Michael

| Production: NX12; Development: VB, TCL/TK, FORTRAN; Testing: NX1847 | engelke engineering art GmbH, Germany | 2019-08-23
| 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)
| Credo: Provide help if you want to get help!

Re: Solid Bolt Pre-load in cylindrical cutting plane needed

Siemens Phenom Siemens Phenom
Siemens Phenom

What does the mesh on the pink region look like?

 

The mesh split does not have to be planar.

 

Re: Solid Bolt Pre-load in cylindrical cutting plane needed

Phenom
Phenom

JimB,

thank you for answering.

The meshes are parabolic hexahedrons.

The connection is done by means of "Mesh Mating Condition => Glue-Coincident"

 

 

2019-07-02_080730.png

 

If you see another way to transfer the pre-load splitting face from planar face inside pink mesh to cylindrical face between purple and pink mesh, that would be great!!!

 

 

2019-07-02_082926.png

 

I assume the problem is the organization of "Which element belongs to which side of splitted mesh?" 

By means of orientation vector X of global CS in my case there is no chance to subdivide the mesh in outer sub-mesh and inner sub-mesh (e.g. according to the radial distance of elemental center point).

 

If there would be the opportunity to define the pre-load by means of a cylindrical co-ordinate system, that would be great.

 

Another way could be if the pre-selection of elements for sub-meshs could be done by user not by program.

 

I upload my model with both cylinder pre-load definitions:

- original and

$*  NX Load and Constraint: B-PL_1kN
BOLTFOR        61.0000+6   23243
BOLT       23243       2       0       1 1025488 1025489 1025490 1025491+
+        1025492    THRU 1025707                                        +
+        1025715    THRU 1025729                                        +
+        1025744    THRU 1025800                                        +
+        1035088    THRU 1035325                                        +
+        1035333    THRU 1035387

- modified.

$*  NX Load and Constraint: B-PL_1kN
BOLTFOR        61.0000+6   23243
BOLT       23243       2       0       1       1       2       3       4+
+              5    THRU     691

 

The model is made with NX10.

 

The inner connecting grid points are renumbered from 1 to 691:

 

2019-07-02_090109.png

 

They describe the connecting area of (E.1= 40) * (E.2 = 5) parabolic hexahedron faces

NoGP = (2*E.1+1)*(E.2+1)+(E.1+1)*(E.2) = 81*6 + 41*5 = 691

 

Maybe I did a mistake by modifying the BOLT definition?

 

If you could find a solution that would be super!

Best wishes, Michael

| Production: NX12; Development: VB, TCL/TK, FORTRAN; Testing: NX1847 | engelke engineering art GmbH, Germany | 2019-08-23
| 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)
| Credo: Provide help if you want to get help!