You'd have to use finite elemet methods to assess the stress concentration effect around the hole. You'd then have to consider the effect on the fatigue life of the component given that you have cracks already around the hole from the manufacture. ESDR might help with factors for edge effects but having already seen failures from bolt holes too near the edge of a plate you're better sticking to the code guidelines should anything go wrong.

corus

Corus

Any small software or shareware you know that does the FE modelling faster.

In addition I am talking about purlins where fractures are not likely.

respects
IJR

Hi IJR

I think you should stick to the code.
A minimum value is just that if their was a tolerance either way then the 1.5D wouldn't be a minimum.

Regards

Desertfox

Desertfox

There are cases when you are squeezed for space, and almost no alternative than to go deeper into the world around the 1.5D

In my professional experience, that and similar situations, is where you employ your engineering capabilities. I have gone through some situations like that where you leave the code and go into papers that were used to develop the code.

One thing about US and Eurocodes, is that US codes come with rich commentaries, whereas Eurocodes come supported by rich documents published by special groups of code experts. So you can around the code.

This is rarely possible in most other codes.

That is the base of my discussion

respects
IJR

IJR - for structural steel connections governed by AISC, Table J3.4 (p16.1-63 3rd Ed LRFD) gives Minimum Edge Distances.  Note that footnote [a] provides a caveat saying "lesser edge distances are permitted to be used provided Equations from Section J3.10, as appropriate, are satisfied."  The equations being referred to are bolt bearing and tearout equations.  Therefore using this caveat if you have a low force level you can use less and less edge distance.  Note that in the direction of load you will eventually reach the code minimum limit of a 10kip reaction.  However currently there is no absolute minimum edge distance - for instance the edge distance perpendicular to a load can theoretically be taken as 0 according to the code as there is no load in this direction.  Obviously this is not wise nor practical and an absolute minimum in the range of 1 bolt diameter from the center of the hole is likely to be added in a future code revision.  

I think the min thickness is based on
a simulated cone models projecting to
the opposite side.  Thicker might not
be an advantage if this is true
from a clamping standpoint at the inter
faces.  I thought there were guidelines
for different values than 1.5 depending
on different thicknesses of the individual
parts.

"Minimum" is a unilateral tolerance.