UK Timber codes

[Wopshistos]

Help! can anyone clarify the correct effective length to be adopted for compression in the design of a traditional rafter with tiling battens ( house roof say)?
Full lateral restraint can be used for bending but Ley=0.75L or 0.01bL yields incompatible results with published tables.
For example a 38mm wide rafter will have i=11mm which yields a max eff. length (180i) of 1974mm but span tables go up to 5.8m for 225mm deep members.
Invariably deflection criteria govern the max. design length of such members but full calculations seem to fall foul in my opinion if these effective length formulae are used.
Are there alternatives, say Ley<=30b for instance???

 

[pba]

I've not got the code to hand but I'll try to work from memory.

A traditional rafter is held in place vertically by the wall plate only and horizontally by the ceiling tie and an equal and opposite force system at the ridge board. The bending stress is caused (principally) by gravity load from roof coverings etc acting normal to the rafter section. The axial load in the rafter is caused by the component of load acting axially. - I take it that is what you are talking about? Purlin roofs are different.

Bucking about the minor axis is normally considered to be prevented by the connection of the tiling battens so your unrestrained lengths are different for the two axes. In the major axis the length is that between ridge and ceiling tie. In the minor axis the length is the spacing of the tiling battens - say 300 - 600mm.

For your example of a 38x225 minor axis slenderness is (say) 450mm/11mm = 41. Major axis slenderness is 5800mm/65mm = 89. Both less than the limit of 180.

NOW - For the tiling battens to act as restraints they have to do two things:
1 - They have to be able to resist the force imposed on them.
2 - They have to be able to transfer that force to somewhere where it can be resited (load path).

Tiling battens vary in size but are usually considered to be 19x38 (see BS 648) - By inspection these are likly to satisfy the requirement 1. Requirement 2 is much harder to prove. I would guess that the force is dissipated over several adjacent rafters. Note however that this assumption is explicitly forbidden for the design of steel framing...

 

[pba]

By the way - 38x225 is probably not a common traditional rafter size. Cut roofs are usually made of 'chunky' timber members - Personally I tend to shy away from timber less than 44mm wide but I have no good reasoning for this.

 

[Wopshistos]

Thx for the reply. I agree entirely with what you have said but this really is a matter of interpretation of the code, BS5268 parts 3 and 7.5.

There is no clear distinction made about the axis of buckling but the code does say that battens provide effective restraint for bending only. There are then other formulae given which I cant see how they are used - particularly in the appended span tables.

Working things out backwards doesnt' help much as nearly all rafters fail under serviceability deflection. Effective lengths used for permissible axial stress calculations cant be worked out but it is clear that Ley is never over 30b which is still generous for batten spacings.
I would still like to know what the code means to help with a little programming.

 

[pba]

OK - I still don't have the code to hand...

Part 2 is the general design code.
Part 3 is trussed rafters.
Part 7.5 I assume is the recommendatons for load tables?

I don't see the relevance of any of part 3 as this relates to trussed rafters, which are clearly different.

I don't recall seeing anyting in part 2 which determines what can be used for lateral restraint so assume that this comes from part 7.5?

You say there is no clear distinction made about the axis of buckling. What are you looking for? Strut buckling mechanisms are explained in any number of text books. A strut will buckle about the axis with the lowest slenderness. The issue is in determining the slenderness values.

You are suggesting an arbitrary limit of 30b. Which in effect is a limiting minor axis slenderness. I don't see this as any more valid that assuming the battens provide the restraint. You are still (possibly)working outside the code.

I've never attempted to create a load/span table but it should be theoretically possible to duplicate one from the formulae. Have you atempted this? Maybe you could impose your 30b limit and see if the esults are effected?

If all else fails TRADA (

http://www.trada.co.uk)

now publish the old building regulation load/span tables - Speak to them about the validity of their design assumptions!

 

[Wopshistos]

part 2 (Timber)2.11.11 - "With roof trusses, purlins or tiling battens may be taken as providing effective lateral restraints, provided they are adequately fastened tothe top chord and are carried back to effective bracing or other support."[as you have said]

Part 7.5 (Rafters){SCOPE}. . . " it is assumed that the tiling battens are capable of providing lateral load distribution and lateral support."[so far so good]

Part 7.5 (Rafters)"The most severe combination of compressive and bending stresses is taken as occurring at the centre of the lower portion of the rafter and the effective length used (Le) is the full length of this portion."[ooh- no tiling battens then?? assume Lex]

Part 7.5 (Rafters) "The effective length Le for buckling calculations should be taken as L/2 for the uniform imposed
load and long term load conditions,"[continuous rafter section over purlin - 2 bays max]
"The effective length Le for buckling calculations should be taken as 3/4L for the uniform imposed load and long term load conditions,"[Lower portion of continuous rafter]
[so are we only ever talking about lex?? ley restraint should be obvious after all???]

"Lateral support to be provided as part2" [relates to h/b ratio which is 30b for h/b=4 - a standard requirement which should be satisfied by the battens n'est pas?][ i am loosely in the code but looking a bit desperate]

Don't have part 3 (trussed rafters) to hand but this states a limit of 0.01b(lbay) to take account of batten restraint in buckling.[ie. not the distance between battens after all. The value becomes increasingly worse if the member thickness is increased tho. Trussed rafters do behave pretty much the same in compression dont they? the latter rule cant be applied as ley as it ensures rafters seem to fail in compression which is hardly ever the case. Deflection controls.

Still in the dark.