Efficiency, getting real data

[Oran]

Thanks, I’ll do my best as I’m also determined to get results. I’m just hoping I can to get the accuracy needed to detect small differences in the efficiency.

[Oran]

Just a quick update on progress. At the moment it is one of those projects that whenever I get a spare moment I do a bit more to it.

Currently I’m measuring the mechanical power output of the drill. The accuracy of the rig is not as good as I had hoped so I’m having to repeat the tests multiple times to get lots of data. This enables me to fit a trendline to the data and convert it to equations. The drills efficiency is at best 55% so I have only been able to get a maximum of 320 watts mechanical power out of it so far.

Next is the interesting bit as I add the NuVinci hub to the rig.

[Normand_Nadon]

The problem with the drill is that it uses AC power on a brushed motor with a "dimmer". The efficiency curve will never be linear and will change as the drill heats up... So to get steady data, you will have to monitor the drill's temperature, speed, current etc on the control experiment and the final experiment...

Keep on the good work Oran!

[Oran]

It’s not so much the drill temperature and the efficiency curve, which is indeed a curve, but the measurements that are the problem. I have to record 3 different readings, rpm, current and drill power input which are all fluctuating slightly. I might experiment with recording a video and then noting down the reading after by watching it frame by frame.

It should still be possible to get reasonably accurate data, it just means collecting a lot more of it. This is something that is normal when measuring the power curve of a wind turbine because it is not possible to record the wind speed at the rotor blades, it has to be somewhere else nearby. The readings don’t match up but by averaging the data a reasonable accuracy can be achieved.

The graph of electrical power input against mechanical power output has turned out well and I now have this equation. Mechanical power output = 0.0004x^2 + 0.3955x - 20.688, where x = electrical power input.


I’m also using the data to calculate and graph the required load resistance for different hub ratio’s so that the power can be kept constant as the rpm of the alternator changes. I will chose from the available resistances and then use the continuously variable nature of the hub to fine tune the power transferred. The test will then be repeated at different amounts of power transfer. I know this means the ratio can’t be at values I might want it to be but if the data is graphed it’s possible to fill in the gaps and find the efficiency at the desired ratio.

[Normand_Nadon]

Sleep deprevation due to a screaming newborn leaves me not understanding what you are talking about ... I will re-read this in a couple of days!

Keep us posted!

Try http://www.imageshack.com/ ... I like it a lot more than photobucket and it is a lot more reliable (personal experience here!, no scientific studies!!!)

[Oran]

Thanks for recommending Imageshack, I’ve just tried it and it seems like a much better site.

It is beginning to get a bit complicated and I’m not the best person at explain things sometimes.
I now have an equation which I was trying to explain at the end of the last post that can be used to plot a graph such as the one below.

Because the input rpm is remaining constant the alternators rpm will vary considerably. I also want to keep the power transferred constant so the load resistance must vary considerably to compensate.

I mounted the bicycle on the rig today and I’m not happy with the connection between the wheel and roller. There seems to be a lot of drag and not much traction. I’ve also calculated that the rpm of the alternator will be too high when in full overdrive. Also the drills speed will need to be geared down more so that it can be run at a higher rpm. 190 rpm just isn’t enough if it’s to put out 300 – 400 watts. Due to all these factors I have decided to build a custom made rig with all chain drives and just the hub without the wheel. It’s going to involve 4 chains because the ratios are more than I could achieve with normal sprockets and 2 chains. I will also have two different ratios between the hub and alternator to help keep the rpm within limits.

[Oran]

I dragged out all my old chainsets, sprockets and wheel hubs I could find today to make a start on a specialist test rig. This is when a habit of hoarding pays off as I think I have enough to get the required gear ratios.
These are the planned ratios so far which result in the drills speed being reduced to about a tenth. The output of the hub is then multiplied by either 8 or 4 so as to keep the alternators rpm between 500 and 1400.

An 18t on the drill driving via a chain a 50t attached to 11t driving via a second chain a 42t on the hubs input.

Then a 42t on the hubs output driving via a third chain a 28t or 14t attached to a 46t driving via a forth chain a 17t on the alternator.

Seems like a lot of chains but it will allow me to put a reasonable amount of power through the hub at the correct rpm.

[Oran]

This is the progress so far with the efficiency test rig. It has involved a lot of modified sprockets and a lot of chains but it’s beginning to take shape. There are two 42t chainrings mounted on 9-spline sprockets, a 42t chainring adapted to fit a disc brake mount, a 46t chainring secured behind a screw on freewheel, two small sprockets attached with bolts, two hubs and a screw on freewheel.

The next thing to do is power up the drill and see if the chains are lined up OK and tensioned. I think the alternator needs to be repositioned as its weight is pulling it out of alignment.

[Normand_Nadon]

WOW!!! You have just surpassed Fallbrook's own testing rigs

[Oran]

Let’s hope I have because I’m getting very close to being able to run some efficiency tests on the hub. I’ve dealt with a few chain alignment issue so they run smoothly now and don’t come off the sprockets. I still have a minor issue with auto shifting to sort and the load resistor needs a larger range so I will get some more heater wire soon and remake that.

I started running a test with the load resister I have and managed 200W through the hub. The initial results for efficiency are looking very good, as in there is only a small variation in the values. The actual values don’t mean anything until I know the efficiency of all the extra chains and bearings. I will have to quantify these extra loses by somehow bypassing the internal working of the hub or by replacing it with a standard rear disc hub, something I don’t happen to have.