August 9th 2014 Measuring the static load at the pedal spindle.

*

Saturday 9th 70F, 21C, blowing a gale, heavy overcast, raining hard. Rain and high winds forecast for most of the day. Might as well call it a rest day. 

Rest day exercises: I have come up with a relatively easy way to physically check the resistance/friction of the rear axles with different 2WD central spacers and/or "sticky" bearing seals. [And yes, I do know the difference between static and dynamic friction.] I'll have to check later to see if the dynamic resistance can be measured with a weight hanging from the pedal axle via a length of standard string.

First I'll put the trike in top gear on the work stand. Then I'll measure the effort required to just start a pedal moving using a simple spring balance. Fisherman's spring balance scales are ideal being both cheap and reasonably accurate. They come with a perfect hook to pull the pedal axle without adding further complication. I shall read the pull required as I gently exert tension on the pedal axle until the drive train starts turning.

The test is easily repeatable as often as desired to check for variations in resistance. [If any.] Provided the spring balance pull is always at right angles to the crank I can avoid the most obvious test error. Removing the wheels will avoid inertial errors. The advantage of a trike is that it doesn't need a rear wheel to support a free-hub. The 2WD trike supports its free-hub and cassette on the rear axle bearings. 

I cannot easily allow for the rider's weight on the axle bearings because that would introduce more errors due to the rolling resistance of the tyres. Rolling resistance is also speed dependent. As is air resistance to a far greater degree. [Let's keep it simple, stupid.]

The obvious downside is having to remove the wheels and outer bearing retaining circlips. Then partially withdraw the axles, drop the cassette and strip the Trykit 2WD free-hub components between tests. At least the new chain is still clean. So this [and a seemingly endless rest day ahead] is a perfect opportunity. Hopefully I can confirm whether my narrower, central spacer mod is only of psychological benefit. 

I have to say that the trike feels quicker now but variations in wind and road incline make it very difficult to be absolutely sure. Now I have an objective way to check. The mind is easily able to fool itself when it comes to small, subjective variations. The multi-billion, global hifi and electronics industry relies on it to push product to the endlessly gullible with far more money than basic common sense.

The same could be said for "racing" push bikes of course. There is very little objective testing of the wild marketing claims for multi thousand dollar/pound/euro machines! For the same money (or less) you can buy a multi-cylinder, multi-geared, high performance motorcycle. One which goes along far faster than any push-bike without any active pedaling by the rider. You are only ever likely to get breathless with excitement!

Motorcycles have far greater complexity than a humble, Chinese mass produced, carbon fiber, weekend warrior's, road hack and its supposedly "weight saving" obscenely overpriced, decorative bling! Simply pouring away half the contents of your carbon-fiber-caged water bottle will save far more weight than the thousands of dollars you just wasted on "upgrading" to this season's "latest and greatest" groupset! [Until the next season's unmissable and unavoidable updates]

Watch this space: I'll be back! First I have to run the gauntlet of the lashing rain to reach my top secret test lab. Just one of the trials and tribulations I suffer in the cause of science. ;-) 

Well, the clue to my hour's experimentation lies in the "string of pearls" in the image above. After repeated tests and axle rebuilds I can confirm that the difference in static friction between the 2 thicknesses of pawl [2mm and 4mm] separator is 11:17 metric nuts.  

My 4lb spring balance was slightly too insensitive to measure the applied loads. So I resorted to a shoe lace and large nuts to measure the required pull. The lace was simply looped over the pedal spindle with the cranks horizontal. I even tried adding the weighted lace to each pedal in turn to remove any doubts. One extra nut [or one less] was enough to make a clear difference in my tests. I spun the chainset repeatedly each time I rebuilt the axle. Just to free up any temporary misalignment or inaccurate seating of the bearings or pawls from each rebuild. 

So, I can safely confirm that the 4mm thick central spacer in the Trykit 2WD free-hub is applying greater end loads to the bearings than the 2mm brass one which I made in the lathe. These same end loads are increasing bearing friction by 11:17. [Increasing from 170g to 260g][Or 6 to 9oz] [Or ~64% greater static friction. The bottom bracket provides its own drag of course but spins freely with the chain dismounted. 

In each case the added weight was enough to produce only about 45 degrees of chainset rotation before coming to a halt. Which suggests that the changing angle of the weight cord [relative to an initial, right angle pull on the crank] was rapidly reducing the applied force to the pedal spindle as the crank slowly dropped. It would also seem static and dynamic friction at very low rpm are very similar indeed. Had it been otherwise the crank drop would have sped up as it descended. In fact the chainset rotated very slowly and smoothly when the weighted cord was hung on the pedal spindle.

If I had a bicycle power meter I could confirm the change in crank power input required to overcome the change in axle resistance. Though I'm not investing the considerable sum required just to amuse myself further. Those who are running a 2WD trike, particularly for racing or time trialling, may want to examine the option of reducing the bearing end loads by using a thinner central spacer between the pawl holders. Or, at the very least, checking their own, inter-axle spacing is to the 4mm spec. Once that hurdle is cleared you can safely ignore all the rest of this nonsense.   

 

I didn't have the patience to refit the four axle, inner facing bearing seals to check the frictional effect of having those back in place.

Cut-away image borrowed from the Trykit website [with permission] The part labels are my own. The left axle pawls are identical but hidden by the freehub body. 

Of course the friction I measured is a worst case scenario. My trike's top gear is 48T x 12T. So it represents a 4:1 increase in axle friction measured at the pedals. My Stronglight cranks are only 160mm c-to-c which slightly exacerbates the difference compared with standard 170-175mm crank lengths.

Can I really feel the difference in axle drag while pedaling at high rpm in a modest gear on the 38T middle chainring? I feel I do but really can't be certain. Many cyclists will spend a fortune trying to reduce all potential losses in their drive train with the latest ceramic bearings. They will then suspend all further disbelief and shell out a second fortune on weight saving measures. While simultaneously ignoring their own weight and fitness levels.

My understanding of cycling efficiency is that the rider's weight is far more important than that of the bicycle itself. I drag a heavy canvas saddlebag around with me everywhere. This itself is always loaded down with a heavy U-lock. I am often carrying a great load of shopping in addition. It may seem strange but I rarely notice the extra weight while riding. Rather, I judge my own climbing performance on the day. It is often a shock having to carry the shopping/sports bag indoors. I wonder why I did not notice the extra weight on all the steep hills I have just climbed.

Only very rarely indeed have I stripped a trike to its bare roadworthy minimum weight. Then headed out for a ride completely intoxicated by the utterly amazing improvement in performance. It cannot be said that I am overweight myself. At least not any more. Riding a trike is a weight and drag handicap anyway compared with most bicycles. Few today would consider riding a bike with sporting pretensions which weighs as much as a lightweight trike. My Trykit weighs about 13kg when stripped of its usual and essential "shopping trolley" junk. Not bad, but racing bikes in its price bracket are considerably lighter.

If reducing bearing drag makes me feel a noticeable improvement then perhaps I don't really need scientific proof. Any more, in fact, than the vast majority of deluded and often overweight cyclists, who think the expense of their toys is a rational use of their disposable funds. But then, what is? A car? A hole in the road into which you throw money? Just to be allowed the right to sit sweating and and fuming in a gridlock traffic jam? While the planet goes to hell and those nasty, idiot and socially inferior cyclists whiz effortlessly by on either side?

Sunday 10th 62-72F, 17-22C, overcast, rather breezy already. It is supposed to gust to 30mph+ for most of the day with rain or showers later. I suppose I should feel myself lucky compared with the UK. Where 2" of rain fell in an hour in Cambridge. With many roads flooded. A trike may be more stable on flat surfaces but has a wider track and a lower bottom bracket. You may not wobble as much as on a bike but you have three wheels to find hidden drain covers and kerbs beneath the murky waters. So an outing on a flooded road is not the most sensible of journeys on either machine. I still think I'd prefer the trike. Since wellington boots with SPD cleats are rather rare one might like to consider polythene bags over the feet just in case.

I left early without my usual coffee and rolls to get some miles in before the wind really picked up. Heading straight into the wind is always a good idea while one is still fresh. I kept extending my route until I had looped all the way around to have the wind on my back for the last leg. It worked too! So the load of shopping had climatic assistance for the last few miles. Lots of broken branches and twigs on the road from yesterday's blow. With even higher winds forecast. The trike still felt faster than before the axle mods. I'm busily rethinking the Shimano 10sp sprocket spacing to match the Campag Ergo 11sp levers. It will have to go back on the stand while I record how well it changes between gears. 32 miles.

Pm. I spent an hour this afternoon stripping the rear axles again. I double checked the distance between the axle inner ends in case I had been mistaken. Even without the Trykit freehub in place there is still no play in the axle bearings when I waggle the wheels at the rims. So I am confident I have not overdone the thinning of the central shim/spacer. Though it should be remembered that each of the half axles is individually mounted in two bearings retained by circlips.

The hollow axles have left a clear mark on the central 4mm spacer.

A slight difference exists in the spin down time of each axle with the wheels mounted. Though I don't believe the small difference is really worth pursuing now that the wheels spin so freely. My Higgins, fitted with Trykit 2WD, has always been exceptionally free running. I was instructed to set up the [adjustable] axles to leave 4mm clearance between them. For some reason my Trykit trike had non-standard [fixed] axle spacing from new. I delayed doing anything about it because I imagined the friction caused by the larger axle bearings on the Trykit trike were causing the problem.  The rear wheels spin for minutes now and it is much easier to spin them up to a high speed in top gear. [48x12]

Having sorted out the axles I turned my attention back to the cassette. I removed the experimental 9sp sprocket spacers and replaced the original 10sp spacers. Gear changes had been much too slow and sloppy with the 9sp spacers. Suggesting that the rear changer was not moving far enough per lever click to move the chain cleanly onto the next sprocket. The cassette locking ring was then tightened on the Trykit freehub to the 40Nm Shimano spec. using a torque wrench. Gear changes are now much crisper than before though still not always perfect. Cable rerouting may offer further progress. I am still struggling with the Ultegra triple front changer. It really doesn't like evenly spaced, chainwheel tooth counts. 48/38/28. I am continuing my online search for capacity specs of the Campag Athena 3x11 triple front changer. [Without any success so far!]

It has been blowing a gale since lunch time but no rain so far. Promised to be even windier tomorrow!

Click on any image for an enlargement.

*