Woven Polytunnels

Model polytunnel - made using triaxial weaving

In theory it's possible to combine the roles of structural support and support for the polythene. This page explores configurations that offer more support for the polythene than conventional designs.

Triaxial weaving is one possibility in this context. Triaxial weave is light, resiliant, strong in all directions, very resistant to shearing forces - and yet still easy to construct.

Model triaxial Polytunnel

I constructed a model triaxial weave polytunnel:

Model polytunnel

I used one of the flimsiest materials I had available to me.

The polytunnel was made with struts of wood measuring 1200mm x 8mm x 1.5mm.

I obtained these struts by dismantling a window blind.

These were joined together in groups of two (vertical struts) or four (horizontal struts).

Close up - showing strut joins and weave detail

Struts were joined together with 6mm internal diameter PVC tubing.

There was no need to cut the struts - except for a few at each end of the structure.


The structure is assembled while lying flat on the ground.

First the left-slanting struts are positioned, then the right-slanting struts are positioned on top of them. Finally, the horizontal struts are woven between them.

Even the polythene (and any insulation) is put in place when the structure is lying flat. It is held at the edges using pegs.

Then the structure is erected. Erection consists of immobilising one edge - and lifting the center of the tunnel slightly.

Then the far edge of the tunnel is progressively moved towards the immobilised edge, keeping it pinned down at all times.

Ropes could also be used to pull on the walls of the tunnel during the erection process.

Internal support

Larger polytunnels would need internal support if they are to resist cross-winds. Such supports are not illustrated here.

Additional structural elements

It is often seen as desirable for polytunnels to have a hexagonal cross-section - i.e. vertical walls, and a somewhat pointed roof. This gives a bit more space at the edges - and reduces the chance of water and snow settling on the roof.

To allow this, ropes could profitably be used internally at the roof ridge - and at either side; at the "eaves" of the structure - to create areas of greater curvature, using tension.

Rope makes sense in this context - since it is very strong, very cheap - and very easy to work with.

Roof ridge

To further reduce the chance of water and snow settling on the roof, a single long tube can be added at the top of the structure, lashed on to the top.


A second layer of polythene is the most obvious way of insulating a transparent polytunnel. To do this, a simple way is to hang some light and cheap horizontal struts from the structure's ridge pole - and then add a second layer of polythene over the top.

Unfortunately, insulation seems likely to increase the persistence of condensation inside the structure - which will reflect off some of the incident light.

Strut angles

Conventional polytunnel designs have few structural elements that traverse the length of the polytunnel.

It seems little strength in that direction is needed - since the ground provides good structural support in that direction.

So - a possibility is distorting the triaxial weave to increase the extent to which the vertical struts are vertical - and decrease the number of horizontal struts.

While some adjustment in that direction may be worthwhile, my perception is that it doesn't make much sense to attempt to distort the structure very far in this direction. The horizontal struts stabilise the structure by weaving it together - and support the polythene - their role is thus suprisingly important.

Rectangular weave

Rectangular (rather than triaxial) woven polytunnels are also possible.

This page illustrates the possible rectangular weave designs.

The end of the polytunnel

One possible way of finishing such tunnels would be to throw in three pentagons - and continue the triaxial weave right to the ground. This might waste a little growing space at either end - but that's probably not a critical concern.

The best way of ending this sort of polytunnel - while wasting as little space as possible and still leaving space for an entrance - is likely to be the subject of future work.

See the triaxial sphere page for more about the issue.

Tim Tyler | Contact | http://hexdome.com/