Cancel
Showing results for 
Search instead for 
Did you mean: 

Random response analysis

N/A

Hello,
I'm looking for someone expert in random response analysis for help, since Nastran
and Femap guides are not very explicative. I try to list some doubts:
-I apply an acceleration load on 4 nodes with a PSD input (in g^2/Hz) with a model
in SI units. When I ask as output acceleration I expect to get again g^2/Hz, right?
And for displacements, forces and stresses which units do I get?
-Is it possible in some way to have a Von Mises stress as output from a random analysis?
-I was asked to calculate differential displacement output between 2 nodes as PSD
by random analysis: is there any way to do that? Is it possible to simply subtract
2 PSD displacement curves (taking absolute value)?
Thanks a lot, and any reference to tutorials or guides is helpful.
Marco
3 REPLIES

Re: Random response analysis

N/A

Hello Marco,
Here are responses to your questions

- I do not recommend that you input your acceleration in units of g^2/Hz. Nastran
does not have the understanding of units and all input has to be in a consistent
unit set. Therefore if you are working in SI units, the acceleration should be defined
in (m/sec^2)^2/Hz. Then all other responses come out in their correct SI units.
I understand that it is more convenient to enter the table in g units. So if you
really want to do that, you should use the scale factor on the RANDPS card to handle
the units conversion. For example if this is your base excitation
Freq (Hz) Accel g^2/Hz
20 5.
50 100.
500 100
2000 5
Then using the conversion that 1 g^2 = 9.8^2 m/sec^2 = 96.04 m/sec^2, you can enter
the TABRAND1 card with values of g^2/Hz and use the scale factor of 96.04 on the
RANDPS card that makes the input in terms of m/sec^2. See the cards below.
RANDPS, 100, 1, 1, 96.04, 0.0, 122
TABRND1, 122, LINEAR, LINEAR, , , , , ,+
+, 20.0, 5., 50.0, 100., 500.0, 100., 2000.0, 5.,+
+, ENDT

The resulting responses, including acceleation will be in SI units. Therefore if
you want acceleation to be in g, you will need to do a conversion. If you are using
NX software, it can do the conversion for you automatically when you are plotting
acceleation. I am not sure what Femap does.

- It is not possible with NX Nastran to compute Von Mises stress for PSD. We are
considering an enhancement for NXN 7 that will compute RMS von mises stress for
PSD. If you are familiar with the NX Response Simulation product - it can compute
RMS von mises and we would use the same method in NXN.
- I think the easiest way to compute the diffential displacement is to use an MPC
equation. Create a dummy node, say node 3, and use it to define the relative x displacement
between node 1 and node 2 by the equation x3 = x2 - x1. In MPC cards this is created
as:

MPC, 100, 3, 1, 1., 1, 1, 1.,+
+, 2, 1, -1.
You can do the same for the y and z directions.

Regards
Mark







"Marco" wrote:
>
>Hello,
>I'm looking for someone expert in random response analysis for help, since
>Nastran
>and Femap guides are not very explicative. I try to list some doubts:
>
>-I apply an acceleration load on 4 nodes with a PSD input (in g^2/Hz) with
>a model
>in SI units. When I ask as output acceleration I expect to get again g^2/Hz,
>right?
>And for displacements, forces and stresses which units do I get?
>
>-Is it possible in some way to have a Von Mises stress as output from a random
>analysis?
>
>-I was asked to calculate differential displacement output between 2 nodes
>as PSD
>by random analysis: is there any way to do that? Is it possible to simply
>subtract
>2 PSD displacement curves (taking absolute value)?
>
>Thanks a lot, and any reference to tutorials or guides is helpful.
>
>Marco

Re: Random response analysis

N/A

Hello Mark,
thanks for your help! Now I understand better the situation. I tried MPC method
for relative disps, although results are not exactly as I expected them, because
far from resonance relative disps should be very little, but they aren't... anyway
at least I have a result to work on.
Thanks again
Marco
"Mark Donley" wrote:
>
>Hello Marco,
>
>Here are responses to your questions
>
>
>- I do not recommend that you input your acceleration in units of g^2/Hz.
>Nastran
>does not have the understanding of units and all input has to be in a consistent
>unit set. Therefore if you are working in SI units, the acceleration should
>be defined
>in (m/sec^2)^2/Hz. Then all other responses come out in their correct SI units.
>
>I understand that it is more convenient to enter the table in g units. So
>if you
>really want to do that, you should use the scale factor on the RANDPS card
>to handle
>the units conversion. For example if this is your base excitation
>
>Freq (Hz) Accel g^2/Hz
>20 5.
>50 100.
>500 100
>2000 5
>
>Then using the conversion that 1 g^2 = 9.8^2 m/sec^2 = 96.04 m/sec^2, you
>can enter
>the TABRAND1 card with values of g^2/Hz and use the scale factor of 96.04
>on the
>RANDPS card that makes the input in terms of m/sec^2. See the cards below.
>
>RANDPS, 100, 1, 1, 96.04, 0.0, 122
>
>TABRND1, 122, LINEAR, LINEAR, , , , , ,+
>
>+, 20.0, 5., 50.0, 100., 500.0, 100., 2000.0, 5.,+
>
>+, ENDT
>
>
>
>The resulting responses, including acceleation will be in SI units. Therefore
>if
>you want acceleation to be in g, you will need to do a conversion. If you
>are using
>NX software, it can do the conversion for you automatically when you are plotting
>acceleation. I am not sure what Femap does.
>
>
>- It is not possible with NX Nastran to compute Von Mises stress for PSD.
>We are
>considering an enhancement for NXN 7 that will compute RMS von mises stress
>for
>PSD. If you are familiar with the NX Response Simulation product - it can
>compute
>RMS von mises and we would use the same method in NXN.
>
>- I think the easiest way to compute the diffential displacement is to use
>an MPC
>equation. Create a dummy node, say node 3, and use it to define the relative
>x displacement
>between node 1 and node 2 by the equation x3 = x2 - x1. In MPC cards this
>is created
>as:
>
>
>MPC, 100, 3, 1, 1., 1, 1, 1.,+
>+, 2, 1, -1.
>
>You can do the same for the y and z directions.
>
>
>Regards
>Mark
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>"Marco" wrote:
>>
>>Hello,
>>I'm looking for someone expert in random response analysis for help, since
>>Nastran
>>and Femap guides are not very explicative. I try to list some doubts:
>>
>>-I apply an acceleration load on 4 nodes with a PSD input (in g^2/Hz) with
>>a model
>>in SI units. When I ask as output acceleration I expect to get again g^2/Hz,
>>right?
>>And for displacements, forces and stresses which units do I get?
>>
>>-Is it possible in some way to have a Von Mises stress as output from a random
>>analysis?
>>
>>-I was asked to calculate differential displacement output between 2 nodes
>>as PSD
>>by random analysis: is there any way to do that? Is it possible to simply
>>subtract
>>2 PSD displacement curves (taking absolute value)?
>>
>>Thanks a lot, and any reference to tutorials or guides is helpful.
>>
>>Marco

>

Re: Random response analysis

N/A

Marco,
I am not sure what the relative displacement should be for your model. But if you
are using the default enforced motion method NASTRAN SYSTEM(422)=1, then you must
remember to use residual vectors PARAM, RESVEC, YES. This could make a difference
in what you are seeing.
Mark





"Marco" wrote:
>
>Hello Mark,
>
>thanks for your help! Now I understand better the situation. I tried MPC method
>for relative disps, although results are not exactly as I expected them, because
>far from resonance relative disps should be very little, but they aren't...
>anyway
>at least I have a result to work on.
>
>Thanks again
>
>Marco
>
>"Mark Donley" wrote:
>>
>>Hello Marco,
>>
>>Here are responses to your questions
>>
>>
>>- I do not recommend that you input your acceleration in units of g^2/Hz.
>>Nastran
>>does not have the understanding of units and all input has to be in a consistent
>>unit set. Therefore if you are working in SI units, the acceleration should
>>be defined
>>in (m/sec^2)^2/Hz. Then all other responses come out in their correct SI

>units.
>>
>>I understand that it is more convenient to enter the table in g units. So
>>if you
>>really want to do that, you should use the scale factor on the RANDPS card
>>to handle
>>the units conversion. For example if this is your base excitation
>>
>>Freq (Hz) Accel g^2/Hz
>>20 5.
>>50 100.
>>500 100
>>2000 5
>>
>>Then using the conversion that 1 g^2 = 9.8^2 m/sec^2 = 96.04 m/sec^2, you
>>can enter
>>the TABRAND1 card with values of g^2/Hz and use the scale factor of 96.04
>>on the
>>RANDPS card that makes the input in terms of m/sec^2. See the cards below.
>>
>>RANDPS, 100, 1, 1, 96.04, 0.0, 122
>>
>>TABRND1, 122, LINEAR, LINEAR, , , , , ,+
>>
>>+, 20.0, 5., 50.0, 100., 500.0, 100., 2000.0, 5.,+
>>
>>+, ENDT
>>
>>
>>
>>The resulting responses, including acceleation will be in SI units. Therefore
>>if
>>you want acceleation to be in g, you will need to do a conversion. If you
>>are using
>>NX software, it can do the conversion for you automatically when you are

>plotting
>>acceleation. I am not sure what Femap does.
>>
>>
>>- It is not possible with NX Nastran to compute Von Mises stress for PSD.
>>We are
>>considering an enhancement for NXN 7 that will compute RMS von mises stress
>>for
>>PSD. If you are familiar with the NX Response Simulation product - it can
>>compute
>>RMS von mises and we would use the same method in NXN.
>>
>>- I think the easiest way to compute the diffential displacement is to use
>>an MPC
>>equation. Create a dummy node, say node 3, and use it to define the relative
>>x displacement
>>between node 1 and node 2 by the equation x3 = x2 - x1. In MPC cards this
>>is created
>>as:
>>
>>
>>MPC, 100, 3, 1, 1., 1, 1, 1.,+
>>+, 2, 1, -1.
>>
>>You can do the same for the y and z directions.
>>
>>
>>Regards
>>Mark
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>>"Marco" wrote:
>>>
>>>Hello,
>>>I'm looking for someone expert in random response analysis for help, since
>>>Nastran
>>>and Femap guides are not very explicative. I try to list some doubts:
>>>
>>>-I apply an acceleration load on 4 nodes with a PSD input (in g^2/Hz) with
>>>a model
>>>in SI units. When I ask as output acceleration I expect to get again g^2/Hz,
>>>right?
>>>And for displacements, forces and stresses which units do I get?
>>>
>>>-Is it possible in some way to have a Von Mises stress as output from a

>random
>>>analysis?
>>>
>>>-I was asked to calculate differential displacement output between 2 nodes
>>>as PSD
>>>by random analysis: is there any way to do that? Is it possible to simply
>>>subtract
>>>2 PSD displacement curves (taking absolute value)?
>>>
>>>Thanks a lot, and any reference to tutorials or guides is helpful.
>>>
>>>Marco

>>
>