Friday, 27 May 2011

What we've been doing at work.

Josh, Caleb, Willis, Joseph, me and Sam

So I've gone here to work on a wind turbine, I should tell you a little more about it.   The design is actually quite a way from being perfected yet, in fact I would say we have only just started to get results which could be useful.   The principle of the access turbine is in its name- it needs to be accessible to those who need it.  That means it needs to be cheap- made locally and out of parts that are affordable to Kenyans.  This puts the expensive high-tech permanent magnet turbines one gets in the West out of the question.  Instead we are proposing to use car alternators, abundant in scrapyards and super cheap.
An alternator is a generator in reverse; the current is generated in the non-moving coil of wire (stator) as the electromagnet spins inside it.  As they stand they don't work very well for us, a car engine rotates much faster than a wind turbine would. But by putting many more coils of copper in the stator, we should be able to get a current at much lower speeds.  This is the theory, and for the first half of my stay the others have been trying to cram in as much copper into the iron core of a stator.  It is a very slow process, weaving the wire in and out of 36 slots- the most recent rewind has had 80 coils in each slot.  Scratching the fragile varnish insulation can make a short circuit, so winding requires a lot of care and frustration.  One can imagine how annoying it is to assemble a rewind that has taken two days, only to find a short circuit which requires an entire disassembly.
the hi-tech wind switch

Meanwhile I started off developing a simple wind activated switch.  Because the rotating part is an electromagnet, it uses power even when not spinning, draining the battery.  This is not a problem in a car as the the whole system is manually switched on (the ignition), but requires our design to have a switch to energise the coil when there is a useful wind blowing.  I had considered all manner of complex ideas to achieve this based on the blades reaching a certain rpm, but a simple flap lifting when the wind picked up sufficed.  Mark 3 was the simplest solution, made predominately out of the steel side panel of an old PC.  Designing things involve walking firstly to the friendly yet small scrap metal yard to see if they have the material in mind, then if they don't, going to the Aladdin's cave of junk where the owner tries his best to rip one off.

  Fabricating the design is done mostly on the dusty floor using our tools- drills, saws, hammers, angle grinder and an arc welder.  Any more advanced process such as turning on a lathe is outsourced to one of many other businesses in the area.  If our local machinist is busy he lets us use his geriatric old lathe.  Pretty much every standard industrial process can be made here, but only Jua Kali fashion- literally translated as 'in the blazing sun' but has become to mean slapdash.
So about halfway through my time here we assembled our latest alternator with the turbine body and blades Sam had created.  It is his first design, improving upon the many previous prototypes created by the other two, none of which had worked.  We took it down to Hippo Point and tested it with a fairly strong breeze coming off Lake Victoria.  Results were poor- the little bades didn't harness the power required to turn the rewound alternator.  Generation at lower speeds requires more physical load (torque).  Not only did we need to make them bigger, we needed to consider a more efficient shape, given that our current ones were mere triangles of a PVC drain pipe.

I proposed sculpting wooden blades, so after a bit of online research I went to the Kibuye timber area and set about getting a sample blade made.   Amongst hills of wood shavings and decrepit old bandsaws lie carpenters' shacks, bustling with men pushing lengths of wood through guard-free planer thicknessers, powered by crackling bare wires.
The wood yard is stacked with wild timber, rough cut before seasoning so all of the lengths are split and warped, and rarely parallel. I find a length of cyprus which looks okay and explain to a carpenter the alien concept of cutting the length of it at a diagonal, changing the angle every 20cm.  He nearly got it right, only messing the last angle at the base of the blade.  I bought some glue and salvaged the last section, and the following day tried some different guys to cut the angle on the other side and plane the aerofoil. I was pleased with the result, but considered their business minds a little bit greedy.  Back opposite the entrance to home I noticed a couple of carpenters sawing away in the shade of a little workshop.  I decided to see if one of them and do a better job, so I introduced myself and got him going.  Willis was keen and meticulous with his tools, and produced another very commendable blade with a hand saw with much more precision than the Kibuye lot.  I had no problem in leaving him to make four more.  I went back to Jua Kali to prepare a mounting and a geared system- an idea Sam had to increase the alternator speed.  I bought some Chinese bike parts and some block bearings, and set about putting them together on a box section frame.
In addition I have been investigating using the first wooden blade shape to form other plastic and fibreglass blades.  I have created a two-part concrete mould from it and we have cut and heated PVC pipe, pressing it into the mould to form two halves.  After that, the mould has been used by a fibre glass specialist to make a test blade, so already we have quite a few blade options to evaluate.  These have just been crude experiments, the wooden shape has to be tested and developed to find the best aerofoil angles.

In order to see which blade performs best, we need to compare the power generated against wind speed. Without a wind tunnel or an anemometer, we took advantage of the slightly more relaxed road traffic laws here and secured a box frame to the roof of the car.

So in the last couple of weeks we have given ourselves many options, and there will be a need for plenty of further testing and development to find out the cheapest, easiest and most effective design.  It is a shame to leave the project before reaching a significant milestone, but I have a strong inkling I will be back.