Hello,

Wondering if someone can explain a simple matter (not for me).

I have four models, (1.5" OD)

1a. Small leg vertically (Y) up (3.5") to a 3" rad. bend that turns into X and continues 15" from the bend tangent intersection point.
1b. Same as 1a but the 15" long pipe is split into 10 elements.


2a. Same as 1, but no bend, only two straight pipes forming the turn.
2b. Same as 2a but the 15" long pipe is split into 10 elements.

The results from a simple modal analysis are as follows:
(Only first mode frequency reported)

Model 1a/845 Hz
Model 1b/970 Hz

Model 2a/1310 Hz
Model 2b/1290 Hz

My questions:

a. Why are there opposing trends between the two types of models (one type has a bend and the other only pipes), i.e., the first frequency increases (845 to 970 Hz) when there is a bend, but decreases (1310 to 1290 Hz) when there is no bend?

b. Also, why is the starting frequency so much different between models 1a and 2a?

c. Specifically, how does the modal analysis treat a bend vs. a pipe in this example?

Btw, lumped mass or consistent mass models did not change the trend.

My client claims that NASTRAN does differently. I do not have the numbers yet from it.

Thank you very much for your input.