Speed trials on our automatic, open source CEB press, The Liberator, are complete. We achieved a maximum brick pressing rate of 16 bricks per minute, or less than 3 seconds per brick. We have run sufficient tests and worked out the bugs in the control scheme, and the results are as optimistic as predicted. After optimization-modifications of the CEB controls and control code, we have gained several bricks per minute in performance – primarily because of running the two hydraulic cylinders contemporaneously, as opposed to sequentially. We predicted this based on initial results from over a week ago. Here we show a half-minute video demonstrating the furious speeds achieved. We are pressing 2.5″x6″x12″ bricks in the first part of the clip, and 3.5″x6″x12″ bricks in the second part – with rates of 16 and 13 bricks per minute, respectively. The power source is LifeTrac at 24 gallons per minute of hydraulic flow:
Read more about these results and control strategy.
First, the brick quality can be better than shown in the clip. It rained 2 days ago, so our soil pile is too wet. The edges show mud that squeezed out the corners of the compression chamber, as the soil was too wet. In general, the brick will be as good as the quality of the soil that one uses.
Regarding the control scheme, the Arduino – the open source micontroller that we are using – is quite flexible. It was relatively straightforward to program simultaneous motion of the two hydraulic cylinders to increase the pressing rate. This works to our advantage because we have a bottleneck in the small hydraulic cylinder for the soil-loading-drawer. We are forced to reduce its hydraulic flow, or it would move too fast. In the simultaneous motion strategy, the main cylinder now siphons the excess flow from the small cylinder, simply via a modification in the control code. This is a trivial change to conceptualize – in retrospect – but it was not obvious that this would be desirable until the actual performance of the machine was observed in field testing.
We had issues with what looked like the valves locking up at high speed, which prevented us from getting full rate results in the last trials. It turned out that it was not the valves that were problematic. Instead, the simple problem was such that at increased cylinder speed, the sensors were moving faster than they could be detected in time, so the cylinders ended up hitting their stops. The solution was moving the detector-magnets closer to the detector, to allow proper detection at higher speed.
The above adjustments – namely of the position-detector-magnets – have to be done by the user upon startup. This has to be done only once, however. Once the machine is adjusted properly, everything works well. These adjustements have to be made because our machine is flexible – and the adjustment compensates for the size of the hydraulic power unit used. If the machine had a dedicated power unit, which it doesn’t – then the settings would all have been adjusted at the Factor e.
Brick uniformity is a requirement for optimized brick laying – which means that one has to use less mortar to lay up a horizontal wall. To address this issue, the machine will operate in two modes. The first mode is uniform brick mode – where brick thickness is adjusted by making the cylinder compress to a certain position. The second mode is timed mode – where timing, instead of position – is used to determine the brick thickness. The former case assures that the bricks are exactly uniform in size. This means that the compressive strength of the brick will vary – if the soil properties change. In the latter case, the timing assures uniform compression, at the expense of brick thickness uniformity – if the soil properties change.
This is a fundamental limit of a vertical pressing machine like ours. We are presently not interested in doubling-to-tripling machine complexity and cost for the sake of building a horizontal-press machine. A horizontal-press machine produces bricks of uniform height, but variable width. An easy solution to assuring that our walls are horizontal is laying bricks on their sides. We will experiment with this and report on the results.
Overall, the progress to date is awesome. 16 bricks per minute is at the limit of physical possibility of a single-brick machine – and we’re glad to report that we have achieved this rate. We may never be able to lay bricks in a wall – straight from the machine – as fast as it’s able to produce them. We may also have to upgrade our soil pulverizer to keep up with this demand for soil. We’ve achieved serious performance, sufficient for an industrial rate of building. Also, it’s good to note that the next machine comparable in performance to ours costs $70k and up. We spent $3500 on materials to produce the present machine, and so can you. More details on fabrication are forthcoming.