Shipping Container Engineering

If you look in our office, there is a whole bookshelf of references on wood frame design.  We have 4 textbooks on wood design.  We have the National Design Standard for Timber Construction, the Wood Frame Construction Manual, both from the American Wood Council.  We have the International Residential Code, which has extensive prescriptive design information on wood.  There are also about 5 different government publications we’ve obtained.  It’s a pretty extensive library.

What do we have on shipping container house design – the only book we could find, a small self-published book by Paul Sawyers.  If you go through the web, there is nothing definitive on shipping containers where engineering is concerned.  We get calls from people all the time who point out this website or that website, but again, no website includes information on how to structurally engineer these things.

The main problem is there is no standard design for the containers.  The ISO standard is a performance spec for the manufacturers.  It gives loads in Kilo Newtons that the container has to carry.  That works well if you are going to make a house from containers that you don’t modify.  What happens when you take off the skin to open the container up?  The lateral bracing disappears, as does a significant amount of it’s ability to carry a vertical load.  So, if you put a bunch of the containers together, how do you know they will carry the wind load, or the live and dead loads imposed?

One way is to figure on the containers having no real strength after they are cut open, and put in a lot of steel to make up for it.  That can be expensive.  You can figure on the corners maintaining the strength to carry the vertical loads (probably a good assumption), and the side rails having no strength (expensive again).  The way we’re trying to go about this is more definitive.  We’ve taken measurements of actual shipping containers and built a model from the measurements.  For the corner posts and side rails, we used steel members that were close (and slightly smaller) than the ones we measured on site.  For the steel skin, we use a steel section modeled from one corrugation and spaced at the same space as the corrugations.  The ends of the members are assumed to be fixed.

Here is a graphic of our structural model from RAM Elements:

I spent a lot of time working on the model and had to make a couple trips to measure and check out containers again and again.  The problem that took some work to get through is the sides and top get a significant compression load during windstorms.  Since I modeled the skin using cold formed members at the same dimensions of the corrugations, and spaced the same, the skin is in effect a series of columns axially and laterally loaded – that includes the roof.

Since the corrugations are connected, and their ends are fixed, there are some column factors that have to be considered.  The columns have to be modeled as braced in one axis, and fixed at the ends.  They also have to be modeled for Continuous Lateral Torsion.

 

 

 

 

 

 

 

 

 

 

 

To the left is how a shipping container deflects (very exaggerated) under a wind and an interior live load.  This container is supported at the connection points.  Look at how the roof ends up in compression, as well as the bottom and sides.

 

 

 

 

 

 

 

 

 

 

 

This is the code check of the model under a 40lb/sf live load in the floor, and a 90 MPH wind load.  I figured that would work, the next step was to see how much wind load I could put on it before it fails.  I tried 24 lbs/sf of wind load on the side, and the container is still way below the allowable stress.  I haven’t run the calcs to see what kind of a wind load that is, but I’m estimating it’s over 120 MPH.

 

 

 

 

 

 

 

 

 

 

In this screen I’ve put a 50 lbs/sf load on the container, this is close to a 200 mph wind load, as you can see, it still doesn’t fail.  HOWEVER, that’s because we haven’t removed the sides.  Once the sides are removed to any extent, the strength drops dramatically, to almost zero.

OK, what happens if you remove the sides?  Such as you want to have a number of containers together and wider open spaces.  So, I tried that with a 40 PSF Live Load for the floor and 20 PSF Roof Load (standard loading for one and two family residences from the 2006 International Residential Code).  The load combination was DL+ 0.75 RL + 0.75 LL (no wind load). Here’s the results:

The red areas are in failure.  As you can see, it fails miserably – the main structural elements no longer work.  So, to make the house structurally safe, I have to go back in and beef everything up.  There are some tricks to this to allow it to be done fairly easily, but you need the ability to weld continuous welds, and you have to have a very skilled welder for this type of work.

A lot of shipping container buildings I’ve seen designed by others do not use the container structure at all – they build a frame around the container.  This is of course intellectually lazy, and it is wasteful of the container.  It also adds unnecessary cost to the construction.

Generally, you need to work with the container structure.  Keep your loads going through the corners, minimize your cantilevers, and keep the amount you cut out of the sides to a minimum.