7 Replies Latest reply on Jan 20, 2015 11:34 AM by Jared Conway

    Stiffness of plate steel for stair treads?

    Gary Garrison

      I am designing a steel staircase and the customer wants a single center stringer design.  The stair treads are to be the "floating" type supported by a small box frame underneath.  My question is this?

      • How thick should the steel plates be to support say 500lbs and not flex excessively assuming the load is at the far extreme end of the unsupported tread?  The unsupported portion of the tread overhangs the "box" support by approximately 14".  I am hoping to be able to remove unnecessary weight by not needing crazy steel thicknesses in the steps.  It's already going to be beast to install as a one piece unit.


      I can't find anything for reference materials that would give me any clear idea of where to start.


      Here are the basics of what I have.

      • 1/4" steel "box" frame approx. 7.5" x 20.5" on the flat top surface.
      • 1/2" steel plate 12" x 48"
      • 1/4"machined aluminum tread
      • center stringer is 8" x 8" x .188" square steel pipe


      There will be no supports underneath the stringer, and most likely a cable or rod baluster system anchoring up through a grid ceiling out of sight.


        • Re: Stiffness of plate steel for stair treads?
          Outcast Studios

          i would say that 1/2 inch is not going to flex much under 500 lbs. you need to look up the deflection of the plate? here is  a simple calculator;


          eFunda: Plate Calculator -- Simply supported rectangular plate with point load


          how are the attachments at the stringer? welds?

          • Re: Stiffness of plate steel for stair treads?
            Bryan Kohn

            Hi Gary,

            You really should find and work with a PE on approve your final design since the consequences of failure could cause property damage, injury or loss of life. This goes double for a commercial application.


            That being said, and for educational purposes only, I ran a simulation of your tread assembly only. I suppressed the "tread angles", suppressed all the holes, and filleted some of the sharp points to make the simulation run easier.


            The first run was the steel plate only and the entire 500 lbs in the corner. The defection is about 4.8mm. A bit high, but probably ok when you consider a 500 lb man standing on one leg just on the corner is going to be a rare occurrence. I looked up stair defections and the values ranged all over the place from L/150 to L/600, which for a 48 wide stair would range from 2mm up to 8mm, so I am not sure what standard you would like on this. I look at stresses for all the cases and that is not a problem. You need to focus on deflection to make sure people feel confident and the stairs don't flex or bounce as folks use them.

            Stair Tread Displacement 500 lbs.JPG

            The next run was the same except including the top aluminum plate. Deflection dropped to 2.7mm.

            Stair Tread with top plate Displacement 500 lbs.JPG

            Considering the load as an edge load which better represents a foot, the deflection is only 2.1 mm.

            Stair Tread with top plate Displacement 500 lbs edge.JPG

            So you are probably good to go on your current design, it will be plenty stiff. But very heavy...


            What happens when we try 3/8" plate to save weight? Looks like back up to 3.5mm.  Not too bad at all and worth considering.

            Stair Tread with top plate Displacement 500 lbs edge 375 pl.JPG

            1/4" plate would save a lot of money and weight but the deflection rises to 6.7mm. That's getting spongy.

            Stair Tread with top plate Displacement 500 lbs edge 250 pl.JPG

            What if  you could talk your client into a small flange? Lets try 1/4" plate with a slim 1" flange on the leading and trialing edges of the flange. only 3.1mm deflection. Better than the 3/8"!

            Stair Tread with top plate Displacement 500 lbs edge 250 pl 1 flange.JPG

            This is just for the tread, but you really need to consider the twist along the long center tube. I did not see how long this this is, but a lot of weight to one side near the center of the stringer could result in a twist or flex.


            Finally, this baby is going to be heavy, coordinate with the other trades and make sure the structure will handle this beast. It is going to weigh as much as a truck.

            • Re: Stiffness of plate steel for stair treads?
              Jared Conway

              are you looking for help on setting up the simulation?

              someone to run the simulation?

              someone to interpret the results?

              or someone to comment from experience on what you should use?


              seems like a pretty straight forward simulation to run as bryan shows

              • Re: Stiffness of plate steel for stair treads?
                Gary Garrison

                We have an engineer what will be looking at this before it's finalized.  It's still good advice, thanks.


                Wow Bryan, you really ran with that.  You confirmed what I was already suspecting.  One thing for sure it will be HEAVY (my latest version is about 2800 pounds).  It wouldn't be so bad of they hadn't requested the square center stringer be rotated 45 degrees with the corners oriented vertically.  That took away my nice and clean (and easy!) mounting surfaces.  It also totally ruined any chance of using a rectangular stringer like my initial design had used a deeper web to handle the load without being so heavy.


                The tread frame to stringer attachment has already been kicked back for redesign.  What they do want, they didn't say... as usual.


                I don't get to run a lot of simulation and typically find that what I need to know is usually one Simulation package higher than what I have subscriptions for.  Consequently, I struggle with setting up and running simulations.  I try to keep up, but things change quickly.  Normally I will only be attempting a simulation once or twice a year.


                I will be moving into more of a product development role instead of a drafter, so I will be doing more FEA.  Mostly what I am trying to do where is come up with as close to an "as built" design as I can.  So to answer Jared's question, from the collective experience of the group, Is what I have typical of designs like this, or is it a case of "I don't know, what I don't know"?