An idea that keeps popping up in my head, that I don't know enough structural/civil engineering to know if it's good from a technical point of view.

Take a piece of polyethelene (PE) piping, fill it with concrete, let the concrete bind. My understanding is that under compression, a concrete (or other) pillar will 'want' to shear apart in a plane 45° to the direction of the compression (assuming we don't bend our pillar). This is an outward movement. This shear force could be contained by the surrounding pipe, since PE is quite good in tension. Othe materials for the pipe might work the same, I say PE because it's quite corrosion resistant.

I know everyone is building concrete pillars with rebar inside, I don't claim my idea is superior. I'm not interested in understanding why everyone is building pillars the way everyone is building them, I want to understand the flaws and limitations in my idea. Some thoughts:

I think it's useful to think as pressure in the pipe, as we have that data easily available - if we use PN10, it can take 10 bar, etc., and we don't need to think about the thickness of the pipe and it's yield strength. But how would the compressive force translate into pressure on the pipe? The concrete is no liquid, so pressue will be less than compressive load / pipe area. How much? I think understanding this will tell us how much, or little, our pillar will carry.

Another possible issue is that PE is quite smooth. The pipe can't take forces via friction and in my mind that translates to point loads at the places where the concrete happens to deform most.


3 Answers 3


Your idea could be ok, but depending on the situation it has a major flaw.

The PE pipe would basically be there to contain the wet concrete, while we wait for the concrete to cure (i.e. to set and strengthen). This is the situation where we need to worry about pressure from the concrete on the pipe. Once the concrete is cured, however, the concrete on its own is completely capable of supporting a lot of vertical force. We design concrete columns based on the compressive strength of concrete; we don't need to add anything else to stop the shearing failure you describe. And the horizontal force on the pipe caused by vertical force on the concrete is unlikely to be an issue.

So, why do we put reinforcement inside concrete columns? The answer is that a column doesn't just take vertical load, it normally takes horizontal forces and bending moments too. And these forces create tension. Concrete is not very good in tension. So we add steel as steel is good in tension.

Can we use the PE pipe to take the tension? I haven't looked into this in detail. In theory, yes. But you'd a) need thick enough PE and b) need some way to transfer stress between the PE and the concrete. As you say, PE is smooth, so without the friction to transfer the stress we're probably not going to meet point b).

So, in conclusion, your PE pipe would work well as something to construct a concrete column which will take vertical compressive loads only. But as currently proposed it's not actually going to add to the load carrying capacity of the column.


Compared to steel, polyethylene (PE) is a weak material. Steel is an order of magnitude stronger than PE.

This site from the US gives a tensile yield strength for PE of 3800-4800 psi, which is 26-33 MPa and this other site from South Africa gives a tensile yield strength of 26 MPa ($N/{mm}^2$ is the same as $MPa$). Whereas this site lists the tensile strengths of various steels used in making pipe as ranging from 48-90 ksi, or 311-620 MPa which significantly stronger than PE.

Prior to using PE as a mould for a pillar it needs to be determined if the PE pipe is strong enough to support the weight & the resultant pressure from the wet concrete just after the pipe has been filled without the pipe deforming, usually by bulging out at the bottom. Once the concrete has set, the pipe is superfluous. In some situations it could be detrimental as the pipe loosens and allows moisture to collect in the gap between the set concrete a the pipe. PE is very susceptible to deformation from heat.

The other issue with PE pipes is that most are not as straight as they need to be to act as moulds for significant concrete pillars.

To give tensile strength to a concrete pillar, a specially designed steel cage is inserted in to the centre of the mould, for the full length of the pillar prior to the mould being filled with concrete.

  • 1
    $\begingroup$ re weight from wet concrete ... PN10, 10 bar can hold about 100m water head, given a density of around 2,5kg/l for concrete that's 40m of 'concrete head'. I doubt your statement that once set, the pipe is superflous but I can't do the math to disprove it. Your other points are well taken. $\endgroup$
    – mart
    Commented Apr 29, 2015 at 11:56
  • $\begingroup$ The statement that the pipe is superfluous once set is almost certainly correct. Among engineering materials, PE has a very low adhesion. Additionally, Andy's answer explains why the shearing failure is not an issue. $\endgroup$
    – regdoug
    Commented May 12, 2015 at 23:00

This approach has been used in many, many scientific studies, and is an active area of research in the materials engineering community, but with a different material. FRP (Fiber Reinforced Plastic) is used to wrap around an existing concrete column (or using an FRP pipe filled with concrete), with very good results.

Sources: 1 2 3 4

  • $\begingroup$ Not sure I buy wrapping existing columns as being particularly similar to the OP's idea, but filling an FRP pipe with concrete would be. Is anything in particular used to bond the concrete to the inside face of the FRP? $\endgroup$
    – AndyT
    Commented Mar 31, 2017 at 8:54
  • $\begingroup$ It's easier to test in the laboratory with wraps. The FRP resin bonds fairly well to any roughened surface. Using a centrifugal cast FRP pipe, you can have the surface roughened on the inside, which then provides an interface area for the concrete. $\endgroup$
    – Mark
    Commented Mar 31, 2017 at 15:21

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.