Sticking Aerogel to steel

[Beau]

18 hours ago, Great_scot_selfbuild said:

The thermal property of tilebacker boards is nowhere near comparable with aerotherm though.

It's not but you can plaster straight onto it. If you are planning to clad this in plasterboard you could use 20mm XPS and have a thinner overall build up.

[JohnMo]

18 hours ago, Great_scot_selfbuild said:

this aerogel or spacetherm?

Same stuff as in the photo by @Mulberry View. 5 years since I bought it.

 

 

[Mulberry View]

20 hours ago, Great_scot_selfbuild said:

Is this aerogel or spacetherm? From the suppliers I’ve approached, aerogel is like a stiff board/sheet (these pictures look like a loose fabric). It appears that the term ’aerogel’ is used across a variety of products, but not all with the same thermal value.

 

What I bought were Spacetherm A1 pads (I'd call them sheets, but they are pads on the invoice). I had a lot of trouble getting consistent advice on what is what. Proctor are fairly helpful, but they still take for granted that products they work with everyday will make sense to DIY'ers like us!

 

It came in an "8x4" sheet, rolled up. It's reasonably flexible, but they encourage the use of separate pieces on my steel column rather than wrapping it round the corner and I think that was good advice. It would follow a curve for sure, but you wouldn't want to wrap it around 2 sides of an object.

 

FYI, I paid £515 plus VAT for 2 full sheets. The carriage was £40+VAT (included in that £515).

 

I'd post you a small offcut if you want to see it before you fork out?

[SteamyTea]

I think the fibre matrix is made from polyester.

So think of it as a very dusty blanket that has been gathering dirt at your grandmother's for the last 60 years.

Then think about what is sticking to it, and why it does not bend tight to make a good hospital corner.

 

Sleep well in this heat everyone 

[saveasteading]

On 08/07/2026 at 18:51, Nickfromwales said:

Just use XPS tile backer boards

Been reading without comment as I don't know aerogel.

 

But my inclination is  to keep it simple and ensure a good, solid and permanent connection.

What area of steel are we talking about, relative to the wall area? i.e. is it the biggest deal to get maximum insulation?

And what depth is available?

There are several makes, too and its available at my local BM (and yours) , Topps and at Wickes, so about £30/m2 and no transport cost.

 

Depending on detail and access, I'd probably stuff the void of the steel to stop that cavity being cold, and also reducing overall heat loss.

 

[SteamyTea]

16 minutes ago, saveasteading said:

reducing overall heat loss.

Not sure but it may be to do with condensation aliveation. 

[Gus Potter]

On 08/07/2026 at 14:55, Great_scot_selfbuild said:

I am planning to fit a layer of 10mm aerogel to the outside of a steel column. Any recommendations on glue / other methods? (Please state if you’ve managed to do this before - my experience of contact adhesive is that it can behave very differently based on the material type and I’m not sure how well aerogel will take to it.

Ok a bit of food for thought.

 

You have a steel column, likely supporting a structural load. Let's say as a minimum you need some fire protection? Say 30 minutes. 

 

Now you can achieve that in two common ways.

 

1/ Box it in with say Gyproc Fireline board. Which probably gives you a detailing problem, which is why you are probably wanting to use Aerogel as it saves space. 

 

2/ Paint the steel with intumescent paint. But for intumescent paint to work it needs to have space to expand into... you see the dilemma? As a rough rule of thumb the intumescent paint thickness needs to expand some 50 times to work properly. It needs space to do this. So you can't stick Areogell to the steel per say and you need manufacturer approval to stick it to intumescent paint. 

 

Now often BC etc don't pick up on this. But if something goes wrong, there is a fire, the steel fails and the building falls on say the Fire Brigade.. then the buck has to stop somewhere.. as an SE I'll be on the radar, Architect's also, and you if you have taken it upon yourself to become a designer then you are facing a huge liability. 

 

Now the above is a worst case.. but if you get a smart BC officer that knows about this stuff then they might be minded to fail your design unless you can prove otherwise. 

 

There are cases where the steel is well protected by masonry and very heavy and thick and thus possibly as at Paulie

On 08/07/2026 at 16:25, Super_Paulie said:

i used a spray contact adhesive on my internal upright, stuck no problem.

I think in Paulie's case the steels are so heavy they don't need fire protection. The load we use when designing steels for fire protection get reduced as they are called an accidental case. In lay terms we don't design most steels for the building being fully loaded up and a fire starting at the same time. 

 

I would go back and look at what your steels are doing, the loads and so on. Also have a chat with your SE  to check if what you are proposing might invalidate their design for example. 

 

I appreciate you may not like this news.. but it's up to you and your risk. 

[Great_scot_selfbuild]

11 hours ago, Gus Potter said:

the steels are so heavy they don't need fire protection

Not that it applies to us, but my interest is piqued - how 'heavy' for fire protection to not be required?

 

(thanks for the other points - we pored over the fire protection discussions early on in the design stage with the PD, SE and TF company; I've since filled my head with MVHR, UFH.... and need to go back and re-read it all.)

[saveasteading]

It's one of these area over perimeter things. So a chunky steel possibly needs no protection, an certainly less... I've had lots of cases, and knowing the principle can save money and also be easier to build.

(To the extent of using a structurally inefficient steel because of the fire and geometry benefit... this seldom happens in a linear design process. )

 

It's the exposed area, so with a fire on the inner face, only that area counts, divided by the whole cross section.

Thus a column in a wall is less exposed than one in an open area. 

 

The info is provided by manufacturers of protection products. 

 

 

[SteamyTea]

1 hour ago, saveasteading said:

It's the exposed area, so with a fire on the inner face, only that area counts, divided by the whole cross section

Can you explain a bit more.

I assume it is exposed surface area that needs to be reduced or protected.

[saveasteading]

2 hours ago, SteamyTea said:

explain a bit more.

Without referring to documents..

It takes a huge temperature rise for steel to soften and fail.

If it was all exposed to fire the heat would rise more quickly than if some is protected.

A skinny section of steel conducts the heat more quickly.

So a typical example of partial protection is being partly enclosed by a wall or having a concrete plank floor sitting on it.

 

The steel is not being protected in the long term. Continued heat will make it fail. But it is being designed to remain structural for a period of time.

 

Thus we could find that an exposed light column needs lots of intumescent paint while a heavy section needs nothing or a thinner coat.

 

I've dealt with 3 fire damaged steel buildings. An agricultural one was a write-off.

One we had built a few years earlier  only needed cleaning and painting.

The third was within an external fire wall (steel cladding  and fibreglass insulaion) ..  no damage other than to the paint on cladding and column   (fire officer amazed).(the fibreglass was converted back to  sand).

 

I don't know if that's a complete or logical answer, so keep asking.

[Nickfromwales]

1 hour ago, saveasteading said:

Without referring to documents..

It takes a huge temperature rise for steel to soften and fail.

If it was all exposed to fire the heat would rise more quickly than if some is protected.

A skinny section of steel conducts the heat more quickly.

So a typical example of partial protection is being partly enclosed by a wall or having a concrete plank floor sitting on it.

 

The steel is not being protected in the long term. Continued heat will make it fail. But it is being designed to remain structural for a period of time.

 

Thus we could find that an exposed light column needs lots of intumescent paint while a heavy section needs nothing or a thinner coat.

 

I've dealt with 3 fire damaged steel buildings. An agricultural one was a write-off.

One we had built a few years earlier  only needed cleaning and painting.

The third was within an external fire wall (steel cladding  and fibreglass insulaion) ..  no damage other than to the paint on cladding and column   (fire officer amazed).(the fibreglass was converted back to  sand).

 

I don't know if that's a complete or logical answer, so keep asking.

For a box section of equal dimension, this I assume would be difficult for heat to affect easily, but I beams (UB & UC) with different thickness materials making the one profile, thus reacting differently over time / temp, the focus would be more on those?

 

However, the ‘system’ would just blanket the solution for worst case, methinks, vs giving chapter and verse for each different scenario.

 

Just my uneducated guess…..

[SteamyTea]

Until the yield point is reached, it is really just differential thermal expansion (in a fire situation) that causes the buckling point to be lower. So what may be fine at 50°C, may have problems at 75°C even though it is well below the critical temperatures of the material, due to the change is sizes (expansion) in 3 dimensions.

This can be made worse if 2 of the 3 dimensions are fixed in place.

Think of a beam that is rigidly fixed at each end and has a weighty slab on it. When it heats up, there is only two degrees of freedom, one downwards and the other sideways.

A bad design would not take those small movements into account, a good design would, and know which way it would ultimately fail.

Another way to think of it is removing a few bricks from the base of a tall chimney. Each individual brick does not carry much load, so removing it cannot do any harm!

[saveasteading]

3 hours ago, SteamyTea said:

only two degrees of freedom, one downwards and the other sideways.

That's another issue than the fire protection, but a chunky beam that is inefficient in normal use can then use that extra stiffness. 

Next time you're in a big warehouse , reasonably modern, if you look up at the rafters, they are likely to have diagonal struts. They're not for fire but show how stiffness can be provided.

Likewise if a floor is built in line with a beam, or a wall into a column  it won't distort.

[Gus Potter]

 

On 11/07/2026 at 08:39, Great_scot_selfbuild said:

Not that it applies to us, but my interest is piqued - how 'heavy' for fire protection to not be required?

It depends! But here is a rough outline of how we go about designing steel fire protection.   

On 11/07/2026 at 11:31, SteamyTea said:

Can you explain a bit more

The following is in the context of the self builder, not multi storey, multiple occupancy structures.  

On 11/07/2026 at 15:51, SteamyTea said:

Until the yield point is reached

Ok, the process / theory is.. roughly this.  

The starting point is to understand how steel behaves when it gets hot. Someone ask me / others a while back.. why does my oven not start to melt / distort when I cook a pizza hot? 

 

The above is what we call a fire curve (SCI, Steel Construction Institute data ). There are different fire curves. They vary depending on fire loading (how much and what types of stuff can contribute to a fire, how "intense and rapid" the heat delivery) but the above is one we would refer to in a typical domestic self build.  

You can see that up to about 350 Celsius the steel maintains much of its strength, which is why your oven does not "melt" and fall to bits. After that the steel strength starts to plummet.  

When we fire protect steels as @saveasteading says

On 11/07/2026 at 14:16, saveasteading said:

 

The steel is not being protected in the long term. Continued heat will make it fail. But it is being designed to remain structural for a period of time.

This is correct. All we are doing it to stop the steel from getting hot enough, then soften too much, and thus not carry the loads on it for the time required by the building regulations for example.  

 

Now we can either select a geometry of steel member and make it very heavy such that there is so much mass of steel that it does not get heated quickly enough to soften to the point of failure.

In lay terms. If we had a solid round bar of steel say 150mm in diameter it has a large cross section area compared with the exposed surface. This is what we call the heated perimeter / cross section area ratio (Hp/A),  the Euro codes have slightly different way of presenting but they mean the same. They are presented using the ratio Area / Volume, A/V.

 

But if we have the same cross section area of steel and weight in an I shape (a universal beam) it has a much larger surface area and thus will heat up much more quickly and thus lose strength more quickly.  

Now we ain't going to be using solid 150mm diameter steels. What we could look to do is is insulate the steel. This could be by way of fire proof plasterboard or intumescent paint. But also we could build the steel into a masonry wall and maybe have only one side exposed to the fire on the inside of say a garage. Or we could have the steel built into a floor that in itself insulates some of the exposed faces of the steel.  

 

Below is a screen shot of a table that shows how we need to consider the exposed sides. They key here is that the lower the section factor the slower the steel will heat up.  The section factor changes depending on how many sides of the steel are exposed.

 

 

But to go back to what @Great_scot_selfbuild asked., which was how heavy does a steel need to be so it does not need fire protection? The info below is taken from Corus literature 2003 as an easy generic example.

 

 The bottom line  (Hp 61m^-1) is of interest to us here in a domestic self build. Using a design temperature of 550 deg you can see that the bottom blue line hangs in there, all other things being equal for 30 min. 

So let go back now and find a steel section with a section factor of lower than 61 m^-1 to try and get a handle on how big a column needs to be for example. 

A 203 x 203 x 86kg/m is a candidate if it has three sides protection (section factor 60), but that is a heavy steel. You also need to handle this safely on site.  Ok let's see how a box section compares. 

 Ok we can get down to 60 for three sides exposed protection.. not much difference. But if we reduce the sides exposed the numbers get a lot better. I've not shown them here. 

 

That main remaining key bit is what we call the load ratio mentioned above as 0.6 ratio.

 

When we design a beam / columns or pretty much any structural member in your house we design for the normal expected loads. The self weight, permanent loads and the imposed ( live loads, people, book shelves etc) but in a fire we recognise that the floors, roof etc are probably not going to be fully loaded with people for example during a fire, the permanent / self weight  load is still there. So we can make a reduction in the load if a fire occurs. Its based on probability and the loads that could be reasonably expected in the case of a fire. We call this an accidental load case and the safety factors get reduced for example. If we designed everything for the worst case buildings would be too expensive to build! The load ratio is the load (stress generated) during a fire compared with the ultimate capacity of the member.

 

Often we design beams to be restrained by say a timber floor.. but if this has burnt away we can't count that. 

 

To conclude:

 

The above I hope, gives you all a bit of insight into what we need to do / think about to design steels for fire. The take away is that sometimes it works using a heavy steel to avoid the expense of fire protection. The design can become much more complex if  bolted / welded connections need to be taken into account. 

This kind of design consideration is a bit complex if you've not done it before, it's not common bedtime reading!

 

It can be a tricky subject once you get into detail, something that many Architect's,  BC officers are also not too familiar with. But it's a fundamental part of producing a safe design. My objective in writing is to try and help BH folk gain a bit more knowledge so you can ask and phrase any questions you have to your designers from a lay person perspective. 

On 11/07/2026 at 15:51, SteamyTea said:

Think of a beam that is rigidly fixed at each end and has a weighty slab on it. When it heats up, there is only two degrees of freedom, one downwards and the other sideways.

You make a good point. Yes steels expand in a fire and will change shape and distort the surrounding structure, to the point it often has to be later demolished. The objective is that the building hangs in there so it does not fall on the fire brigade, set light or fall onto a neighbouring building not least. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Edited 11 hours ago by Gus Potter
Formatting.

[saveasteading]

On 11/07/2026 at 15:51, SteamyTea said:

only two degrees of freedom, 

Next time you are in a big steel warehouse, look at the rafters. They will be very slender if well designed. To prevent rotation/ out of balance of the rafter, there can be angled struts from the bottom flanges to purlins.. a small amount of steel adding much strength.

 

350°C , as mentioned by Gus, is very hot indeed. By the time that temperature is reached everyone is gone a long time ago, and most contents will be too.  Windows will fall out, doors will fail, skylights will burn or break and so some heat will escape. Ouside walls may even fall apart or burn. Or if it stays intact the oxygen demand doesn't keep up with the fire.

 

The fire bbrigade'sprimary mission is to save life and that is long before 350.

But also the fire must not spread so boundary walls and compartment walls are to a higher spec, and steel columns must stay in place.

That moves us onto pinned or fixed bases which may be for lesson 2 (only if requested).

[SteamyTea]

1 hour ago, saveasteading said:

lesson 2

Yes please.

[saveasteading]

OK. 

Fire first. If a building near a boundary is on fire then it must not be allowed to spread to adjacent buildings or materials. We assume an empty plot will be built on.

So the wall on that boundary must stay standing even if the rest of the building is collapsing and trying to push or pull it over.

So, let's stick with steel portal frames.

The turning forces in a collapsing building are huge and we have a very chunky column fixed down very robustly. Beneath that the foundation also has to be over a big area to resist turning.

That is a fixed base.

Again to confuse the staff in the warehouse by staring.. you will see that the column is chunkier on that wall than on one facing the car  park.

On a gable wall everything is lighter.

 

 

Where it isn't a fire wall, the portal column can be skinnier at the base, with few bolts and the foundation is also much smaller, only taking vertical loading. That is a pinned base. Not usually literally pinned but sometimes they are.

 

On a 30m span shed the boundary foundations might have 5m3 of concrete each, and mustn't cross the boundary,  while the others have 2m3 or so.  it's all money.

 

 

My business used tapered columns (and rafters) for best value, and they were thin at the bottom to deep at the top, except on fire walls. The depth of section mirrored the forces.

 

The fire wall cladding is fire rated, and the columns usually need to be protected too.

 

That's all a bit messy and bitty but I hope makes sense.