Are non axial loads to columns correct??

[Nick Jackson]

Hi Tony,

It is possible that non axial loads to columns are not being calculated correctly.

Please see attached email.

Regards

Nick J

[Tony Bryer]

Re: Are non axial loads to columns correct??

I haven't had the email. Are you referrring to steel or timber columns? Note that the load offset is from the face of the member so the resulting moment varies with the member size.

[Nick Jackson]

Re: Are non axial loads to columns correct??

Hi Tony,

I have found a file extension that allows the file to be attached - (please try getting a word document (.doc) under 19.5K !!) - Bigger limits would help !!

I finally got it attached as a zipped pdf file.

The problem is to do with load offsets on a capped steel column (Superbeam and Prosteel affected) If you set up a 75kN load on face 4 of a 102 x 203mm UB it comes up with a Myy moment of -0.20 (Superbeam) and -0.27 (Prosteel). I am using an offset of 0 for a capped column. For this section I would expect 0.051 x 75 = 3.825kNm for Myy - or am I reading BS449 and BS5950 incorrectly?

If I manually add the moment, the required size of column increases dramaticaly as you might expect. This may an issue if the bug is confirmed as columns may have been under specified.

Regards

Nick J

[Tony Bryer]

Re: Are non axial loads to columns correct??

Thanks, I haven't looked at the zip but with load positions 3 & 4 on an I-section beam the offset is from the face of the web so an offset of 0 will give a moment of load x web thickness/2. To get what I think you are looking for you would need to manually calculate the offset = (flange width - web thickness)/2.

[Nick Jackson]

Re: Are non axial loads to columns correct??

Hi Tony,

That explains your rather low Myy values.

However, I think working from the width of the web may be incorrect. I have attached an extract from BS5950 which states that the moment for a capped column should be calculated from the face of the column, not the face of the web. See 4.7.7 sub para 1).

This makes a certain kind of sense as the supported beam will deflect slighly placing all the load (theoreticaly) on the edge of the cap plate, or face of the column.

Your method works fine for a beam bolted to the web of course. But then your help files state that the offset for a capped column should be 0 where it should be width/2 as BS5950 requires you to start measuring from the face of the column.

To fix this issue unequivocably, you would need to offer the user 3 choices:
1) Case - beam on cap plate on column, Myy=F x width/2
2) Case - rafter on cap plate on column, Myy=0 (warn user to check connection does not impart significant moments (whatever significant means!!)
3) Case - all other cases, use 100mm from face of column or centre of stiff bearing whichever is the worse.

I have had a quick play - if you select a 100 x 200 RHS section, the moment is correct - Myy=3.82kNm

So in essence, this all boils down to where the face of the member/column is taken from in the case of an I Beam - either the face of the web or the face of the cap plate.

Regards

Nick J

[Nick Jackson]

Re: Are non axial loads to columns correct??

Hi Tony,

Prosteel user guide arrived yesterday and contains the 3 column cap conditions that I listed above so no need to tell you this.

This issue boils down to where the face of the member is taken from, the face of the web or the face of the line connecting the flanges.

Please let me know your thoughts.

Cheers

Nick J

[Tony Bryer]

Re: Are non axial loads to columns correct??

Apologies for the delayed reply: have only just returned from a trip back to the UK. The program works as designed: measuring the minor axis offset from the face of the web is correct when a beam comes into it on a cleat etc. What is wrong is that the documentation says to use an offset of 0 for cap connections: this is correct for load positions 1 & 2 (major axis bending) but not 3 & 4 with I sections. On these the offset would need to be calculated manually using the selected section dimensions. I will correct this in the help file for the next release and when the manual is next revised.