SE refuses to spec beam with underplate

[flanagaj]

Our SE has specified paired beams for the garage and large patio sliders.  I have gone back and asked if we could use a beam with underplate as it makes the brickies job easier.  I got a rather interesting response, which I'm finding rather confusing.

 

"

Is it possible to change the spec of the beams for the garage and large patio doors, to instead use beams with underplate for the garage and patio doors.  The reason I ask is that we are not having a soldier course and it will be the brickwork tricky to marry into the outer leaf beam if I use paired beams. I get the problem, however, if you just use a plate, there’s a theoretical possibility of instability of the lintel, as the combined effect of weight of the brickwork and the leverage to overturn the beam  is greater than the available restraint loading on the blockwork side. 

See attached sketch showing what I mean.

So I’m not too keen on this option in this case

 

However,  I can’t see a reason why not to  use an angle  in the outer leaf; the 200x100x12 RA should be sufficient, in both cases.  Any good?"

 

Said drawing I was sent is attached.  Given the ends of the lintel are supported, I'm confused as to how it could physically twist like she is indicating.

 

 

PATIO AND GARAGE LINTELS.pdf

[Mike]

I'd guess that there's a risk of the plate sagging and pulling the internal lintel over with it. The proposed solution seems entirely reasonable - SE's do tend to be very practical people.

[flanagaj]

16 minutes ago, Mike said:

I'd guess that there's a risk of the plate sagging and pulling the internal lintel over with it. The proposed solution seems entirely reasonable - SE's do tend to be very practical people.

Ok, thanks.  The proposed solution will make the outer leaf brickwork easier too

[Gus Potter]

2 hours ago, flanagaj said:

Our SE has specified paired beams for the garage and large patio sliders.  I have gone back and asked if we could use a beam with underplate as it makes the brickies job easier.  I got a rather interesting response, which I'm finding rather confusing.

 

"

Is it possible to change the spec of the beams for the garage and large patio doors, to instead use beams with underplate for the garage and patio doors.  The reason I ask is that we are not having a soldier course and it will be the brickwork tricky to marry into the outer leaf beam if I use paired beams. I get the problem, however, if you just use a plate, there’s a theoretical possibility of instability of the lintel, as the combined effect of weight of the brickwork and the leverage to overturn the beam  is greater than the available restraint loading on the blockwork side. 

See attached sketch showing what I mean.

So I’m not too keen on this option in this case

 

However,  I can’t see a reason why not to  use an angle  in the outer leaf; the 200x100x12 RA should be sufficient, in both cases.  Any good?"

 

Said drawing I was sent is attached.  Given the ends of the lintel are supported, I'm confused as to how it could physically twist like she is indicating.

Your SE is trying to convey to you some very difficult / complex structural behaviour, in lay terms. To put this into context. Working from left to right. The first two diagrams show a Cee section. It takes about 3-5 years after graduating from uni to even get you head round this, to be able to design something buildable and cost effective.  Your SE has made a pretty good go of the diagrams.

 

 

 

 

Option 3 works, up to a point span wise, much depends on the span / size of the angle. Angle sections are prone to twisting. Your SE later may want to also restrain the angle in places to stop it twisting. 

 

Below is a cross section though something very similar to what you have. 

 

 

 

Buried within the insulation you'll see two bolt heads. below is a screen shot of the steel GA details.

 

If you look closely you'll see a steel fin plate in the middle of the angle span. Here I'm shortening the effective length of the angle and stop is twisting over the doors below. The fin plate method is a small local thermal bridge, but an effective way of getting a relatively small steel angle to work. Yes there are still thermal bridges, but this is not a new build rather a retro fit into an older house. Below is the proposed rear elevation. The clear opening width is some 4.0m. The steels over the opening need to carry first floor and roof loads. I've only posted parts of some of the drawings, hope enough to give you flavour of some of the things we need to think about. 

 

 

[LnP]

Is it the three fins (labeled F4) which prevent the twisting, or the seven pieces of angle (labeled B8)? What are the B8s fixed to?

[Gus Potter]

43 minutes ago, LnP said:

Is it the three fins (labeled F4) which prevent the twisting, or the seven pieces of angle (labeled B8)? What are the B8s fixed to?

Good spot, thanks for looking at my stuff. There are three fin plates, one at each end and on in the middle marked F4. The end ones stop the angle rocking on the external leaf of masonry. The centre one halves the effective length of the steel angle. Over a 4.0m length the angle will fail, in torsion and buckling. 

 

The members marked 7 xB8 are the first floor joists that rest on the inner leaf beam. 

 

I've added more drawing to let you see what I've been up to! Remember thought these are screenshots of part of my drawings. While I'm happy to share some stuff I do I need to protect my intellectual property to some extent. It is after all my day job also. The stuff you see is only part of what is a much larger drawing set. 

 

This is big bug bear I have. Architect's nine times out of ten don't put a grid on there drawings at the concept design stage. Thus it makes it right from the get go hard for a Client to line stuff up. As I wear two hats.. I just can't see why they don't do this, it's bonkers or arrogance? 

 

Edited Friday at 23:26 by Gus Potter

[ETC]

I’d say it doesn’t cross their minds to put a grid on.

[flanagaj]

@Gus Potterthanks for explaining things.  That is a complex steel structure for a door opening.

[Gus Potter]

On 15/08/2026 at 07:00, flanagaj said:

That is a complex steel structure for a door opening.

It is. The reason behind it is that the building loses a lot of sideways stability. The steel work is also designed to stop the building falling over sideways. This is often the main challenge as an SE, not just to hold up the vertical loads but also to maintain sideways (horizontal stability). You also need to maintain your thermal breaks / mitigate thermal bridging etc. This makes the SE work harder as in the old days we just tied everything together!

 

That all has to come together and then needs to be detailed (just as important), with the propping sequence thought through,  so it's buildable by an average (hopefully) competent builder at a sensible cost. I deliberately put lots of detail on my drawings where I want to explain to the builder what is expected of them and thus by inference what they should include in their price. 

 

 

[Gus Potter]

On 15/08/2026 at 04:17, ETC said:

I’d say it doesn’t cross their minds to put a grid on.

This is really a bugbear I have with folk who are less experienced Architects, it's even more of a wtf are you doing with Architects who are set in their ways. I wear two hats, SE and Architectural designer. Design is about communication of information not least.  

 

At concept design stage if you stick on a grid it lets everyone see how the floors and things line up. On BH most folk don't have access to say CAD or more complex design packages where we can overlay floors etc.  It lets us all see what goes where, get a rough handle on things and communicate well.

 

If there is a grid we can say, say on teams, go to grid intersection 2A, and a bit to the left or 300m up on the drawing, there is no dubiety and that stop wasting time trying to identify what part of the drawings we are looking at, it avoids error and folk getting frustrated. 

 

Later when we come to the detailed design a grid is essential to allow the builder to set things out. Why not start with a grid at the beginning, grids help reduce error!

 

Now Architect's like their drawings to look good, when I draw stuff in cad I'll change line weights, emphasise at time depending on whether I'm trying to show a Planner, a checking Engineer or how the kitchen may look to the Client.

 

Funnily I work with quite a few highly experienced Architects, they have embraced my suggestion for putting grids on drawings! It's the future!

 

Architects.. For pure visual presentation, just turn the grid layer off and stop being so precious.