Jan Ridders is a model-engine builder of some renown who came across this machine on YouTube. And you can see why it would have caught his attention:
When he tried to build a Thermo Pulse Mobile (his name) of his own though, he ran into some difficulty. Until he tried something that shouldn’t have made any difference whatsoever:
After a lot of random experiments, which I shall not enumerate here because of their irrelevance, I almost gave up until I took a closer look at the video. I noticed a kind of restriction at the place were the brass cylinder was fixed in the glass tube… Because I hadn’t the slightest idea how the restriction should look, I used numerous small brass cylindrical plugs from my scrap box, including those with and without all kinds of bores in them and other deformities. I started random experiments, putting all kinds of plugs in the glass tube at the place where it is fixed in the stand…
By far, most of the plugs didn’t bring any positive effect, but with some of them the engine started to show some sign of life. I was pleasantly surprised and at the same time extremely astonished when I found one plug that caused the engine to keep running, although yet not fully reliable. Using the typical characteristics of this “reference plug” I made a new plug, adding step by step changes to the geometries of the reference plug. This was still a random way of working, but at least somewhat more systematic. Finally I succeed in making a restriction that allowed the engine to run reliably.
The design Ridders ended up with works, sure, but neither he nor his fellow model engineers have been able to explain its thermodynamics. They did, though, come up with a diagram if you’d like to try your own: