Negative contact pressures. CPRESS negative in tension ? 1

[mvp23]

Hi,

I have a working model of 3 concentric cylinders with the cylinder in the middle being the master surface and the other two being the slave surfaces. I have a pic attached of the model , which would better represent my question.

A gap opens up between the middle cylinder and the outer cylinder due to the tensile stresses as evident from the radial stress plot in the pic. In view of this I was expecting to see positive contact pressures on the outer tube . However Abaqus tells me the CPRESS values on the outer tube are negative.

So is CPRESS negative in tension and positive in compression ? I'm just trying to confirm if the CPRESS behavior is consistent with my radial stress distribution results.

Thanks

 

[corus]

No, you can't have negative contact pressure. If you do see negative values then they tend to be very small and are probably due to rounding errors or such like. I tend to modify the contours to show values from 0 upwards.

Tara

http://tinyurl.com/4ydjg7m

 

[mvp23]

I can relate to what you are saying. So does open up , the contact pressure on the slave surface will be zero since no contact is being made. Is that right ?

I'm seeing contact pressures of the order of -4E06 Pa to -2E09 Pa

Thanks for your help.

 

[corus]

It may help to plot COPEN to see where the contact surfaces separate.

Tara

http://tinyurl.com/4ydjg7m

 

[mvp23]

I did plot COPEN for the outer tube and my COPEN values are of the order of 8e-10 meters and corresponding contact pressures of the order of -3E06Pa . In another model when I had a separation of 4e-06 meters I got a zero contact pressure.

Is this because , the value 2e-11 meters is so small that Abaqus still considers it to be in contact ? I could view the separation only when I used a deformation scale factor of 20 ( scale factor of 1 being normal ).

I've attached how my assembly is setup and a plot of COPEN and CPRESS plotted against the angle theta.

 

[mvp23]

and here is the COPEN plot .

Thanks

 

[corus]

Unless I'm reading it wrong, the COPEN graph appears to show negative open values, and where there is a positive COPEN value then you have positive contact pressure. All of which are contradictory to what you'd expect. Values f 1^-10 don't give much confidence in the solution though. Perhaps a change of units to give more realistic numbers?

Tara

http://tinyurl.com/4ydjg7m

 

[mvp23]

From 0 to 240 degrees I have positive COPEN values(vertical axis on the right has the location of its origin slightly different from that of the primary axis on the left ) and negative contact pressures. The COPEN graph does show negative COPEN values at certain locations in the innertube-outertube contact region. This is because the stresses are compressive in this region. Due to the unconventional shape of my middle tube , I get this behaviour. I cross checked this with a middle tube being a perfect cylinder.

Also , I have this file attached that talks about the subroutine DLOAD and states that

" *DLOAD defines pressure (i.e. compressive normal stress)
to be positive, so we apply tensile loading by making the pressure negative."

Does this statement lend any validity to my results ? I mean , I'm seeing separation in most regions , which implies tension and my contact pressures turn out to be negative in the region of separation as well.

Thanks for all the help once again.

 

[mvp23]

Thought I'll attach a separate plot of COPEN .

 

[corus]

If there is no contact then there can be no contact pressure. The only way I can think you have a negative contact pressure is if you don't let the two surfaces separate after initial contact, ie. they are stuck together once contact has been made. If you allow separation then you can only have a postiive contact pressure, ie. a compressive normal/radial stress.

Tara

http://tinyurl.com/4ydjg7m

 

[mvp23]

Thanks for your help. Under normal behavior , I had unchecked the 'allow separation after contact ' box. That was causing a negative contact pressure.