Observatory has no roof!!!

[MarkH]

I knew that,  was trying to cheer you up with a bit of humour. Seriously though, I think if you deflect the prevailing wind you should be ok.  Diffusers may also help.

[Rocket Pooch]

Did anyone read the link I put up there about roof design and wind?

[MarkS]

Yes, it was very interesting.  It indicated that a 15 degree pitched roof is "optimal" for generating huge amounts of lift - easily enough to lift a 650kg roof wholesale.

My son did aeronautical engineering so I plan to solve the Bernouilli equation for that design.

[Tony G]

Mark,

With regards to your Avatar.......you don't wanna do it like that! 
Seriously, it the costs covered on your insurance or are you sorting it out yourself?

Tony G

[MarkS]

With regards to your Avatar.......you don't wanna do it like that! 

I shouldn't do what like that?

I'll be sorting it out myself once the weather improves i.e. probably wait until the Spring - it'll cost a few hundred quid I think

[Ian]

have fun with computational fluid dynamics

btw, IIRC the Bernoulli equation may not be applicable in this case as it applies to a single flow, the walls of the obs separate the flow outside with the stationary air within. However, CFD is full of assumptions and approximations when looking at real situations, this link might be worth a read too http://www.clayroof.co.uk/wind-uplift-calculations.htm

Chris, it's a shame that paper hasn't got some maths to back it up. I don't think it's necessarily as simple as that, particularly as increasing the roof pitch also increases shear loading but it's in interesting read.

I guess another thing to think about is the direction of the prevailing wind on the site and align the ridge as close as you can with it. Or how about a ring of leylandii right around as a wind break

[MarkS]

have fun with computational fluid dynamics

btw, IIRC the Bernoulli equation may not be applicable in this case as it applies to a single flow, the walls of the obs separate the flow outside with the stationary air within.

You might be right.  I'll read through his lecture notes in any case, for a bit of fun.

[Ian]

of course, if you want to go the other way, you can use Navier-Stokes. The wikipedia page is pretty much maths porn...

[Rocket Pooch]

Chris, it's a shame that paper hasn't got some maths to back it up. I don't think it's necessarily as simple as that, particularly as increasing the roof pitch also increases shear loading but it's in interesting read.

Agreed would be better, I think Mark has nailed the next roof anyway the lower one with the tarp stayed on.

[MarkS]

I've done quite a bit of internet research on this and the results are quite counterintuitive.  In terms of negative pressure (i.e. lift) generated by wind passing over a roof, wind tunnel experiments show that wind approaching a gable roof from a direction of 45deg i.e. SW in my case is the worst possible case - it will tend to "peel away" upwards from the leeward front edge.  This probably explains why the snapped 3"x2" timber was on the leeward side of the observatory.

So will a monoslope (pent roof) fare any better?  The consensus from wind tunnel experiments and from the ASCE building code (American Society of Civil Engineers) is that a monoslope roof generates more lift than a gable roof. A flat roof is better than a gable but it is not a huge difference (as indicated by the graphs in the link that Chris posted).

Best roof of all is a steep (30deg or more) gable roof or even better, a steep hip roof.

In general, the biggest negative pressures are at the corners and edges (including the top ridge), so the roof panels need to be firmly screwed to the trusses in those areas and the trusses need to be firmly anchored to the walls.

I found a useful online calculator at:
http://engexp.com/calculators/asce_comparison.cfm

It calculates the negative pressures that various zones of a given roof with a given slope have to withstand for a given wind speed to meet the required ASCE standard.  In pounds/sq ft!

[Fay]

Just got up & think i may have a little read of your suggestions, Ian, with a cup of tea.

[RobertM]

I've done quite a bit of internet research on this and the results are quite counterintuitive.  In terms of negative pressure (i.e. lift) generated by wind passing over a roof, wind tunnel experiments show that wind approaching a gable roof from a direction of 45deg i.e. SW in my case is the worst possible case - it will tend to "peel away" upwards from the leeward front edge.  This probably explains why the snapped 3"x2" timber was on the leeward side of the observatory.

So will a monoslope (pent roof) fare any better?  The consensus from wind tunnel experiments and from the ASCE building code (American Society of Civil Engineers) is that a monoslope roof generates more lift than a gable roof. A flat roof is better than a gable but it is not a huge difference (as indicated by the graphs in the link that Chris posted).

Best roof of all is a steep (30deg or more) gable roof or even better, a steep hip roof.

In general, the biggest negative pressures are at the corners and edges (including the top ridge), so the roof panels need to be firmly screwed to the trusses in those areas and the trusses need to be firmly anchored to the walls.

I found a useful online calculator at:
http://engexp.com/calculators/asce_comparison.cfm

It calculates the negative pressures that various zones of a given roof with a given slope have to withstand for a given wind speed to meet the required ASCE standard.  In pounds/sq ft!

So from all this research, can I assume that it just needs to be fastened down better in the Mk 2 version ?

Might also be worth reducing any overhang for the wind to get under.

Robert

[Ivor]

Did any of the research discuss other methods for breaking the airflow other than the ridge tile? Having a 30 degree+ pitch plus additional strategy may help avoid future problems.

[MarkS]

So from all this research, can I assume that it just needs to be fastened down better in the Mk 2 version ?

Might also be worth reducing any overhang for the wind to get under.

Yes - that seems to be the conclusion.  You are also right that overhangs are subject to the greatest wind pressure of any roof part so I'll reduce them as far as possible.

[MarkS]

Did any of the research discuss other methods for breaking the airflow other than the ridge tile? Having a 30 degree+ pitch plus additional strategy may help avoid future problems.

There are weird things that can be added to the corners to prevent damaging wind vortices (vortices which can increase lift) but I don't think I'll go to those extremes.  A 30 degree pitch would increase height and therefore require planning permission.  I think it would look unsightly in any case.