Gravity in SimXpert

[JLKDS]

Hello guys,

Having a bit of trouble with gravity on an I-beam section on SimXpert. When using gravity under global LBCs, into 'Acceleration vector scale factor', I enter 9810mm/s^2 as my units are in mm, correct? I have checked my density is also in kg/mm^3 (i.e. 7.85e-6 for mild steel). Yet when I run the results, the forces of the FBD are several orders of magnitude too great. I have four point loads over the 10m span of 1500N each and they appear negligable compared to this massive gravitational force so that the reaction forces are pretty much constant on each of the 10 supports.

From the FBD, the gravitational force on each node reads 15.32 kN (Apart from where there is a point load and at the supports there is nothing applied). The figure is in the correct range as the beam has a mass of 15.7kg/m but the magnitude is far too high.

Any thoughts?

Many thanks,

Pete

The model setup:

A section of the FBD:

 

[IDS]

If your beam self weight is 15.7 kg/m and the span is 10 m then the expected reaction for the internal supports is of the order of 1.54 kN, so the result you are getting is out by 1 order of magnitude.

- Have you double-checked all the input?
- What are the maximum bending moments at internal supports and mid-span?
- What are the numbers in the second image, and what do the arrows represent? Are you applying the dead load twice perhaps?

I suggest running without the point loads until you have found the error.

You might also consider running in a simple continuous beam program to get the correct "exact" results.



Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[GregLocock]

If your basic units (MLT) are kg mm s then your force unit is mN

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[rb1957]

is that a mass density or a weight density ?

 

[IDS]

Reading again, I see the 1.532E4 is the node force due to self weight, with nodes at 1 m centres, not the support reaction. As Greg noted, if the accelleration due to gravity is given in mm/s^2 then the force units will be mN (the force required to accellerate 1 kg at 1 mm/s^2), so the calculated forces due to self weight are 15.32 N per node, not 15.32 kN.

So reduce your g value by a factor of 1000 and work in N (your E values will then be in N/mm^2, and numerically equal to MPa units, which is convenient).

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[rb1957]

i think you can enter g in m/sec^2, you need to be careful mixing units, but density must be kg/mm^3.

i think greg and doug have the answer ... mN (kg.mm/sec^2)

maybe next time model in meters ?

 

[GregLocock]

FWIW cars are modelled in tonne mm s so our force units are N

!!!!!!!!!!!!!!!!!!!!!!!!!!!

For my own use I always use kg m s, for the reasons you are running into. This means stresses are stupid big numbers, of course, and deflections are very small ones.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?