A PVC bladed type of helical wind turbine, inspired by the 3 bladed version of TyPower2. In an attempt to emulate the Savonius profile, I used a heat gun to curve the blade tips.
To produce usable power one needs:
- a turbine, wind or water.
- a generator, coupled to the turbine.
- a rectifier to convert the ac output from the generator to dc for battery charging.
- a storage battery, usually lead acid deep cycle.
- a charge controller to prevent overcharging which kills battery life.
This instructable deals with the turbine construction, which will be matched to an optimum generator( no idea yet).
The rectifier will be 2 x Full Wave Silicon Bridge Rectifier 35A 400V, more than adequate, but if you have access to a 3 phase rectifier, so much the better.
I havent had much success with diy pcb's, so I'm going the veroboard route using DIYLC v3 for the component layout software.
Step 1: The PVC Blades
The design began with an idea of 1 metre wide turbine and evolved from there. Using 40mm dia pvc tube, I would need 30 blades to achieve my initial 180 degree twist, which meant the height would be in the 1.4m range.
I sliced the 30 0.5m sections in half and used a pvc joiner for cost reasons, Id only need 15m of tube instead of 30m using other cut methods. First a jig was built to see how feasible it was to cut pvc length wise, it worked well enough and so each 1/2m section was marked prior to the tedious cutting job.
When making 4 marks on the tube the 2nd set of marks always seemed to rotate in relation to the first set, and so I devised an easier method. I made 2 marks on the tube 180 degrees apart and on a table, leaned a ruler against the tube and then drew a line down the tube.
Main tools used here were the jigsaw, a small desktop drill press to make the perpendicular axle blade mount holes, and a lathe to cut out the ridge inside the pvc join fitting which prevented the tube from sliding right through, I wanted a bit of overlap to make a better blade joint.
Step 2: Blade Tip Bending Jig
This was my method of bending the tips into an approximate ideal Savonius shape using a heat gun and plywood jig. Hindsight has shown that it would have been a better idea to have had a full length jig to clamp down the main section while the tip was being bent, but it seemed quicker at the time to have a dowel for length indication in the back.
A dry test fit will highlight if any blades touch each other due to too much tip twist, keep the heat gun handy to rectify twists or warps.
When heat forming keep a cool wet cloth handy to "set" the hot bend with a quick wipe when you have it at the desired shape, it also protects your left hand that holds down the tube while forming it.
Main tool used here was the heatgun on low power which was sufficient to turn the pvc into soggy spaghetti, the forming temp indicator.
Step 3: Holding It Together
The frame has to be resistant to twisting to prevent the binding of the upper and lower bearings due to misalignment, as well as provide a housing for the bearings. The vertical and bottom arm are 25mm x 50mm x 1.6mm rectangle steel, the top was a scrap 25mm x 38mm that was at hand.
For the bearing recesses, I cut a slice out of a steel tube, squeezed it shut around the bearing and slid it into the hole in the ends of the cross member arms.
The bearings are 6000zz type with a 26mm od, the tubing wall thickness was 1.2mm which required a 29mm bi-metal holesaw. A thinner slice of tubing creates a bearing seat stop.
The 6000zz has a static load rating of +/- 198kg and being a deep vee ball bearing, can accomodate an axial load of 99kg, my blade assembly weighing 13.5kg is therefore well within the design limits.
The corner bracing was provided by 200mm sections of 19mmx 3mm thick flat bar.
Main tools used were the welder, small angle grinder and a 29mm bi-metal holesaw in the drill press.
Step 4: The Charge Controller
As I mentioned in the intro, I'll be using the Mdavis19 555 charge controller. In the link to step 8 he is still using his first version with the LM339 chip, but his website has the full 555 version details including fine tuning and test point voltages.
I've made a few substitutes, a 1455 instead of the 555, a 330nF 100v instead of the 0.33uF 35v and a 2N3904 instead of the 2N2222, and the circuit worked first time for me as well. The set points for charge and dump at the test points are on the Mdavis19 website, 1.667v for the lower charge setting and 3.33v for the upper dump setting.
I've included my veroboard version with a DIYLCv3 software screenprint containing the components that I used, mostly because that was what was available, as well as the *.diy project file. In the project file you will be able to see all the cut tracks indicated by a red dot square combo.
Be aware that the software doesnt have a way to view the copper track side in reverse, so I marked all the cut-tracks with a fineliner on the component side and then flipped the actual board to do the cutting.
My scrap veroboard was 14 holes by 32 tracks, hence my particular layout, but its not necessarily the best way to do it.
Main tools were a side cutter and soldering iron.
Step 5: Some Odds and Ends
I used baby bottle teats as weather protection for the top and bottom bearings, packed with grease for added protection.
I used the large flange rivet head but found it undesirable with curved pvc, it tended to distort the blade as it was being riveted, my partial fix was to slice a section off to lessen link strip distortion. I was distracted by finding the optimum position of the rivets to lessen blade distortion and failed to check the blade to blade distance(50mm) and as a result created a 110 degree twist instead of the 180 degrees I initially wanted.
Once the blade halves have been joined they need to be individually balanced before being assembled as a group onto the main axle. A screwdriver in a bench vise will do for the axle. I used a pruning cutter to cut the edges of the heavy side until each blade sat fairly level.
The main axle is 12mm OD with a 10mm ID, the 6000zz bearing also has a 10mm ID, and so I used a polished 10mm OD thick wall ss tube slid through the bearing and into the 12mm main axle as a means of making the bearings accessible for replacement if necessary. Cross drilling and cotter pins mechanically link the bearing shaft and main axle, the blade group assembly is also mechanically coupled with individual pop riveted link strips.
The main tools were the drill press for the cotter pin holes and a plier type pop riveter.
At this stage I have no idea in mind for the generator section, but I have it in mind to take some torque vs rpm measurements once the wind picks up from its current gnat burp and update this 'ible accordingly.