I made a beam

[SteamyTea]

Tomorrow I go back to work after 2 months off.

Now I have totally wasted that time, so thought I better catch up with a little experiment I had in mind.

With some old 9mm OSB, I made a 200mm deep beam.

It is 2.26 m long, and held together with cheap PVA 'no nails' type adhesive (cost a quid a tube).

The flange is 100mm wide.

One flange is 2 layers of OSB, the other is 2 strips of timber, 20mm by 35mm holding the other flange in place.

To test it I put a box on it and filled it with water.

It basically only moved 6mm when I had put 52 kg of mass in the middle.

Some of that will be because the whole thing became unstable on the trellises. The result is shown in the last picture.

Going to be plenty strong enough to make a shed from, so may use cheaper, thinner ply and smaller battening timber.

 

I think I will get @Gus Potter and @saveasteading work out the numbers, my moments if gryration were jumping out the way when the box of water fell.

 

 

 

 

 

 

[joe90]

What are you trying to prove? We know I beams work and are very common now (well they did work in my build 🤷‍♂️)

[SteamyTea]

Just now, joe90 said:

What are you trying to prove

Only that I can easily make one very cheaply.  Probably less than a tenner each.

[joe90]

1 minute ago, SteamyTea said:

Only that I can easily make one very cheaply.  Probably less than a tenner each.

Oh I agree but the problem is “standards”, the bought ones are tested to meet demands in the building industry/insurance/BCO. If it’s fir a shed where no standard has to be met then all well and good. I have no doubt I could knock some up if required.

[saveasteading]

Well i am impressed.

I think I'd put some nails in, in case the glue deteriorates with time or temperature.

I don't think there are any rules for garden shed strengths.

Even so, I have delegated beam strength calculation for some decades so am not about to restart now.

What's the worst load? 1m of snow or a person. Your 50kg load is good enough for me.

The decking board will also act as a flange if fixed well enough. 

 

Now if you can get the materials cheap as seconds...you know the boards delivered with corners bashed that should be in a corner of the bm.

 

[SteamyTea]

6 minutes ago, saveasteading said:

What's the worst load? 1m of snow or a person. Your 50kg load is good enough for me.

A person, but maybe a green roof.  Be nice if I could hang my kayak from it if I can get enough height.

I want a shed where I can open up one side (near enough).  That way I can make the door like a garage door and get an extra 1.8 covered area.  That way I can do dusty work but still be out of the sun, or the rain.

 

10 minutes ago, saveasteading said:

The decking board will also act as a flange if fixed well enough. 

I was thinking of using the decking as the flange.  Make it in panels and then just glue and screw together.

Basically all panels will be the same, i.e. 2.4m by 1.2m and just do infill panels to make the roof slope.

[Dave Jones]

put a couple strips of this down the length flitch beam style to strengthen it right up

 

https://www.amazon.co.uk/Metre-Purpose-Perforated-Galvanised-Fixing/dp/B0098ICTEA?th=1

 

 

[SteamyTea]

3 minutes ago, Dave Jones said:

put a couple strips of this down the length flitch beam style to strengthen it right up

While a bit overkill for a shed, I like the idea.

A bit of PU resin and GRP would make something very stable, and strong.

I had to make some carbon fibre beams once.  Not sure what they were ultimately used for, but amazingly rigid and weighed very little.

[FuerteStu]

My concern here is that when you cut OSB thin, it loses its strength, and water resistance. The cut edges need to be resealed. 

 

Also, OSB edges are rubbish for fixings, they tend to blow out and split easily. 

 

That's why manufactured beams are mostly timber topped and bottomed. 

[SteamyTea]

4 minutes ago, FuerteStu said:

My concern here is that when you cut OSB thin, it loses its strength, and water resistance. The cut edges need to be resealed

Was one of my concerns as well.

I did think, if designed as a stand alone beam, rather than as part of a panel, then a strip of WBP ply glued over the flange would help.

 

It was really something I wanted to try out, gut said it should be alright, but there is nothing like actually testing something out. 

I did my apprenticeship making parts for destructive testing machines. Was interesting seeing the machines being tested themselves.

[saveasteading]

1 hour ago, SteamyTea said:

maybe a green roof.

That is very heavy, especially when wet. Then it might snow.

 

Have you thought about making your own RAAC /aircrete planks? Cement  plus soapy water and a whisk.

[Pocster]

4 minutes ago, saveasteading said:

That is very heavy, especially when wet.

I didn’t read why @SteamyTea posted all this as it looked boring . But a certain phrase I’m familiar with caught my eye 

[SteamyTea]

22 minutes ago, saveasteading said:

Have you thought about making your own RAAC /aircrete planks

Used to make aerated composites, was interesting work.

One part was a stone effect shower tray. Can't think why no one has one. The customer never got us to make the bath tub he designed.

 

20 minutes ago, Pocster said:

certain phrase I’m familiar with caught my eye 

Rigid.

 

[Gus Potter]

On 06/09/2023 at 17:09, SteamyTea said:

I think I will get @Gus Potter and @saveasteading work out the numbers, my moments if gryration were jumping out the way when the box of water fell.

Hi Steamy.

 

If you want a quick set of section properties (unverified by calculation) I can do that.. I would cheat a bit in that I would draw the cross section in Auto CAD, and use the MASSPROP function.. this gives the I (second moment of area) values, centre of gravity, radius of gyration etc. Do a sketch of the cross section and I'll try and post the results. From that you could work out the stresses in the outer fibres. To do this you would use the same approach as you would do for a flitch beam... timber and steel combined.. but it would be more complex due to the geometry.. I won't do that for you.

 

You are using two different types of timber that have a different Youngs Modulus.. thus things suddenly get a lot more complicated to do manual calculation.

 

If you fancy it you can expand you experiment to calculate the average Youngs modulus using a true central point load as follows:

 

For a simply supported beam the deflection Delta = PL^3 / 48 E* I

 

Delta = deflection in mm, I = second moment of area mm^4 , E = Youngs modulus N/mm^2 , P = your point load (N) and L (mm) = the length between the supports. Use units of N (Newtons and mm as you units)

 

This rearranges to E = PL^3 / 48* I * Delta

 

Load up the beam increments and measure the deflection mid span. Plot a graph and you should get a fairly good result for E if done carefully and you have a good gauge... a tape measure won't really do.

 

Now you can compare your E value with that for chipboard. Now to be safe you may want to divide your E value by 1.96 (Conservative) based on a 95% normal distributon. You could use a value closer to the codes of dividing by 1.5 ish. This would give the the minimum modulus for a transfer beam... no load sharing etc.

 

For stength design you want to apply a material factor to your timber / chipboard grade stress of about 1.3 thus you take your grade stresses and divide these by 1.3

 

And again to be safe you want to factor up your applied loads by say 1.5 so you multiply your load you have calculated by 1.5.

 

For the strength check you calculate the factored bending moment by using the formula P* L/4 and then calculate the stress using the formula M / Z = stress, Z is the section modulus. You then check that the applied stress is less than the allowable stress.

 

But there is an inherant weakness in these types of beams an that is web buckling over the bearing point so you need to add in vertical web stiffeners over the supports.

 

Have fun and let us know any results of further experiments!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

[SteamyTea]

3 minutes ago, Gus Potter said:

Have fun and let us know any results of further experiments

Thanks. Making, testing and breaking stuff is a lot of fun. Should be taught at school.

I may in the near future, knock up a beam, and a column, which will be more like the thing I want to build the shed from.

I will be drawing them up in my CAD, mainly to work out the least amount of ply sheets I need to buy.

[joe90]

1 hour ago, SteamyTea said:

Making, testing and breaking stuff is a lot of fun.

That does make sense, make a beam and load it till it breaks or buckles, then you have your answer. I didn’t understand one word of @Gus Potter post above unfortunately, but making and breaking it I get 🤔.

[saveasteading]

41 minutes ago, joe90 said:

make a beam and load it till it breaks or buckles

That's why cathedrals took so long to develop, (and suspension bridges):. a few failures followed by major tweaks.

The Romans overdesigned sometimes.

[joe90]

8 hours ago, saveasteading said:

That's why cathedrals took so long to develop,

It fascinates me that back then with no SE input, scaffolding (as we know it) etc they managed to build such huge magnificent buildings, flying buttresses etc 

[saveasteading]

33 minutes ago, joe90 said:

back then with no SE input

There have always been clever people.

I once saw a fascinating OU programme about arches and buttresses, using wooden blocks, and those ancient masons probably used models too, I feel.

And they had scaffolding.

My amazement is not so much the structure, but the geometry, esp the roof tracery.

[Pocster]

3 hours ago, saveasteading said:

There have always been clever people.

That include me then ? ; winging it for a decade with no architect , SE , BCO input ? 🙄😎

[saveasteading]

1 hour ago, Pocster said:

winging it 

Using your inherent understanding of physics and natural forces?

Some wing it and fail.

Others improvise and succeed.

Then there are the rest of us.

[Pocster]

1 minute ago, saveasteading said:

Using your inherent understanding of physics and natural forces?

Some wing it and fail.

Others improvise and succeed.

Then there are the rest of us.

Yeah , yeah - all that bollocks . 

[Onoff]

On 06/09/2023 at 19:55, Pocster said:

I didn’t read why @SteamyTea posted all this as it looked boring . But a certain phrase I’m familiar with caught my eye 

 

Flange

[Gus Potter]

20 hours ago, joe90 said:

That does make sense, make a beam and load it till it breaks or buckles, then you have your answer. I didn’t understand one word of @Gus Potter post above unfortunately, but making and breaking it I get

You have understood most of it. "load it till it breaks or buckles"

 

What the calculalations and figures do is translate that into how you determine a safe design stength and how much it will probably deflect by if you then made this type of beam and built it into your house.

 

When we design and test stuff it costs money so we don't test to destruction right away. We try and gather as much info on its behavoir before we destroy something.

 

If you look at the load span tables for engineered joists there are loads of different figures for loading and span. The manufacturers don't test every combination.

 

The probablitlity bit is to do with how many beams you test, like say a pull out test on a resin anchor. If you only test one then it does not tell you much. Test three and you can get a better idea of the spread of results. Test five and you gain more "confidence" and that lets you then reduce the reduction factor. Just call the standard deviation. This make the design more efficient.

 

If you think about concrete many millions of cube crushing tests have been done over many years so there is lots of confidence about the spread of results you get when using normal concrete.

[joe90]

25 minutes ago, Gus Potter said:

You have understood most of it.

Actually I wish I had studied this stuff when I was younger (I am way to old now), I believe I have an inherent gut feeling about what will hold and what will break (and I tend to over engineer a bit). I just can’t prove with numbers what I believe. I have a fascination in all things S.E. I did actually think I could make my own I beams when I did my build but I had way too much to do besides that otherwise I  would be like @Pocster and never finish 🤷‍♂️🤣