I am putting in a radiant floor to distribute the heat from the solar collectors I recently added.
The floor is going to be the type where you remove the current finish floor, and add runs of PEX pipe on top of the existing subfloor with with plywood spacers between the PEX runs to make a surface level with the top of the PEX tubing. You then place place your finished floor over the spacers. Solar heated water is then circulated in the PEX tubing to heat the floor.
To make this work efficiently with solar heat, it is desirable to be able to use as low temperature water as possible for heating. This makes it desirable to have low thermal resistance in the finished floor, and to have an efficient means to spread the heat from the tubes to the finished floor efficiently.
Aluminum heat spreader plates are often used to spread the heat more efficiently on radiant floors. There are some commercial ones of various types, but, being cheap, I am building my own using sheet aluminum.
There is also an issue of how to install the spreader plates. They typically consist of a flat plate with a groove or trough formed down the centerline that the PEX tubing fits in. You have two choices on how to install this: 1) with the groove facing up, in which case the spreader plate lives right on top of the plywood spacers (just under the finish floor), and 2) with the groove down, in which case the spreader plate is sandwiched between the plywood spacers and the existing subfloor.
Shows the sandwich style radiant floor with alum heat spreaders. This is version 1. With version 2, the alum spreader plate would be between the "sleeper" and "subfloor".
Having heard arguments for both methods, I did a little test to try to figure out which worked best. I took a 6 ft length of PEX tubing, and used the first 2 ft with the grooves up, the 3rd 2 ft 2 ft with the grooves down, and the middle 2 ft with just the plywood spacers (for a baseline to compare with).
I ran hot water through the PEX, and measured the temperatures in each of the three segements at a three points: 1) over the centerline of the PEX, 2) 1.5 inches to the side of the PEX, and 3) 3 inches to the side of the PEX. In all cases, the PEX and spacers were covered with a thin foam underlayment and an 8mm laminate floor.
The "subfloor" is supported by 2X4's on each edge, and is insulated underneath to roughly simulate a real floor.
The spacers that the PEX is placed between and the subfloor are 3/4 plywood. The alum heat spreader is 0.019 sheet alum (alum soffit material). A groove is formed in the alum that fits the PEX tube OD fairly snugly, and a bead of silicone is laid down in the groove before the PEX is put in it to improve heat transfer. The tubing is generic half in PEX (I plan to use PEX-AL-PEX on the actual floor).
The graph shows the temperature profiles for each spreader plate:
Point 3 is on the top of the finished floor, and directly over the PEX tube.
Point 2 is on the top of the finished floor, and 1.5 inches from the PEX tube.
Point 1 is on the top of the finished floor, and 3.0 inches from the PEX tube.
Tambient = 69F, Twater = 92F
The drop in water temp from inlet to outlet was about 0.5F or less -- so all three segments get nearly the same temperature water.
All temperatures measured with and IR surface thermometer.
I was surprised how much difference there was between the two arrangements (groove up and groove down), which I had been lead to believe would be very similar.
It looks to me like the alum spreader plate just under the finished floor, and over the plywood spacers is substantially better in getting heat to the room. That is, for the same water temperature, it is producing finished floor surface temperatures that are 6 to 7F better -- a near 100% improvement in the temperature drop from tube to finished floor.
For solar heating, where it is desirable to use low temperature water, this seems like a very worthwhile gain. I'd like to hear if there are disadvantages to this way of putting the spreader plates in that I don't see??
The test setup. Stackup is: 1) laminate floor, 2) thin foam underlayment, 3) PEX tube between spacers or sleepers, 4) 3/4 plywood subfloor, 5) 2X4's supports on edges. The temperatures were measured using an IR thermometer on the black tape strips at the numbered points.
Detail of the first two feet with the heat spreader right under the foam underlayment. The bit of brown visible to the right and left of the laminate is the top of the alum heat spreader.
Shows the setup with the finish floor removed, and the foam underlayment folded back.
The closest segment is the "high performing" alum just under the finished floor. The next one is wood with no alum spreader, and the top one has the alum spreader sandwiched between the subfloor and spacers -- the grey you see is the groove going over the PEX tube.
Detail of start transition from no spreader plate to spreader plate under spacers.