Bolt Pretension With Temperature

General Questions
1

Hello, I am trying the pretension tutorial below on YouTube.

I set up a similar model with the video and got the same results. Then I read the reaction forces with the surface defined for preload and observed that the given preload value was maintained. Then I wanted to observe that the preload value decreased by defining the temperature in the 2nd step. However, the results were very different than I expected, although the model’s movement was correct, there was an increase in the reaction force. When I set up the same model in other commercial software, there was a decrease in preload as expected.

I uploaded the model here

https://drive.google.com/drive/folders/1pWirAdlU1hKh5tW98Uy13565YE2wRA4L?usp=sharing

2

Sample tutorial link is given below:

3

Reaction force is given below

image
image237×134 1.31 KB

4

I submitted this input file in Abaqus and got the same results: 1000 N in the first step and -3032 N in the second step. Maybe there’s some difference in your setup in that other software.

5

Could you please try to define temperature dependent material properties. Because theoritically, there might be some loss on preload because of thermal expansion.

1 Like
6

Hello,

You are right, there is definitely a difference among the models. I had solved the same model using Ansys Workbench. The behavior I expect is to carry the preloead to the second step and a decrease in the preload with the temperature. What do you think about this topic? In my Ansys model, I give the boundary conditions in the same way in the same way. Could the places I read in the reaction force be wrong?

When I tried the following example with temperature, I encountered similar problems.

https://www.feaclster.com/calculix/cccx_2.18/doc/ccx/input_deck_viewer.php?input_deck=pret3

7

No, I used *SECTION PRINT for the Load surface and *NODE PRINT for the pretension node (in Abaqus) and results are the same so it’s a matter of the setup. I wonder what differs in your Ansys simulation. Can you share some screenshots showing BCs/loads there ?

1 Like
8

Could you share the inp file?

9

Take a look at this post.

Main idea is that the proper way to pretension a bolt is by displacement up to the desired tension and later fixing it in the second step. If not, the solver tries keeping the pretension to the specified value.

By other hand, section print may not be working properly depending on where you are measuring. I would try to have various ways and locations for measuring reactions in your model.

imagen
imagen976×740 43 KB

ezgif-7-a107688bd6

10

Thanks ANYS,
This the correct form of the preload behaviour. There are numerous way of applying pretension. Applying displacement is one of them. But this method requires several runs, if the initial displacement value is far from the target displacement.

1 Like
11

Have you try pretension foce driven first and swithching to fixed diplacement on the second step?

12

Ansys is known for having issues with their bolt preload implementation. I know from a customer that they are looking into solving the discrepancies with respect to Abaqus results. So, I would be very careful to compare apple to oranges in terms of the bolt pretension implementation.

13

I tried this way.I think it’s like you said

*Step
*Static, Solver=Pardiso
**
** Controls ++++++++++++++++++++++++++++++++++++++++++++++++
**
**
** Output frequency ++++++++++++++++++++++++++++++++++++++++
**
*Output, Frequency=1
**
** Boundary conditions +++++++++++++++++++++++++++++++++++++
**
*Boundary, op=New
** Name: Displacement_Rotation-1
*Boundary
Internal-1_Supp_1, 1, 1, 0
Internal-1_Supp_1, 2, 2, 0
Internal-1_Supp_1, 3, 3, 0
** Name: Displacement_Rotation-2
*Boundary
Internal-1_Supp_2, 1, 1, 0
Internal-1_Supp_2, 2, 2, 0
Internal-1_Supp_2, 3, 3, 0
**
** Loads +++++++++++++++++++++++++++++++++++++++++++++++++++
**
*Cload, op=New
*Dload, op=New
** Name: Pre-tension-1
*Cload
28331, 1, 1000
**
** Defined fields ++++++++++++++++++++++++++++++++++++++++++
**
**
** History outputs +++++++++++++++++++++++++++++++++++++++++
**
*Node print, Nset=Internal-1_Load, Totals=Yes, Global=Yes
RF
**
** Field outputs +++++++++++++++++++++++++++++++++++++++++++
**
*Node file
RF, U
*El file
S, E, NOE
**
** End step ++++++++++++++++++++++++++++++++++++++++++++++++
**
*End step
**
** Step-2 ++++++++++++++++++++++++++++++++++++++++++++++++++
**
*Step
*Static, Solver=Pardiso
**
** Controls ++++++++++++++++++++++++++++++++++++++++++++++++
**
**
** Output frequency ++++++++++++++++++++++++++++++++++++++++
**
*Output, Frequency=1
**
** Boundary conditions +++++++++++++++++++++++++++++++++++++
**
*Boundary, op=New
** Name: Displacement_Rotation-1
*Boundary
Internal-1_Supp_1, 1, 1, 0
Internal-1_Supp_1, 2, 2, 0
Internal-1_Supp_1, 3, 3, 0
** Name: Displacement_Rotation-2
*Boundary
Internal-1_Supp_2, 1, 1, 0
Internal-1_Supp_2, 2, 2, 0
Internal-1_Supp_2, 3, 3, 0
**
** Loads +++++++++++++++++++++++++++++++++++++++++++++++++++
**
*Cload, op=New
*Dload, op=New
** Name: Pre-tension-1
*Boundary, Fixed
28331, 1, 1
**
** Defined fields ++++++++++++++++++++++++++++++++++++++++++
**
*Temperature
NALL, 300
*Temperature, op=New
** Name: Temperature-1: Deactivated
**
** History outputs +++++++++++++++++++++++++++++++++++++++++
**
*Node print, Nset=Internal-1_Load, Totals=Yes, Global=Yes
RF
**
** Field outputs +++++++++++++++++++++++++++++++++++++++++++
**
*Node file
RF, U
*El file
S, E, NOE
**
** End step ++++++++++++++++++++++++++++++++++++++++++++++++
**
*End step
14

I share the inp file in the same directory

https://drive.google.com/drive/folders/1pWirAdlU1hKh5tW98Uy13565YE2wRA4L?usp=sharing

15

Ofcourse I use default pretension eleman in ansys. I give a pretension in first step and In second step I lock the pretension in ansys. In addition I have a similar problem in 2d models. Im think that ı can not read reaction forces properly but i have no idea for the reading.

image
image682×532 29.2 KB

16

Your don’t have contacts between the bolts and the plates.

17

Yeah, it’s a compound merged where the heads touch the plates:

merged
merged1046×725 64.2 KB

merged 2
merged 2553×695 52.4 KB

18

Oh sorry.Your are right.
Then the bolt head is pulling the plates when the thermal expansion takes effect. That could explains why reaction force is changing sign. Small preload compared to thermal expansion.
Do not measure reaction where you apply external forces.