Progress:

The final piece of titanium arrived for Mr Mangle and for once the advertised thickness was spot on; the side armour is exactly 8mm thick. Here is what two bots-worth of titanium looks like:



Just two more sheets to clean up and they can all go out for cutting

The new Revolextrix Lipo pack also arrived. It is well made and is 250mAH higher capacity than the old packs but it is wider than advertised so the back of Mr Mangle had to be extended 6mm. Not a big deal but it would have been if the battery arrived AFTER all the titanium was cut!



The battery came in a strange plastic case which has a switch and a LED on the top. they are not connected to anything and I have NFI what they are for. Perhaps its to get around Lipo transport restrictions?
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

Interesting, Wasn't there some rule that the batteries have to be installed in equipment to be posted?
_________________
www.demon50s.com - Minimoto parts
http://www.youtube.com/user/HyzerGlen - Videoooozzz

Maybe its classified as the world's most crappy flashlight. The case is strong & closes securely; if we are lucky, other manufacturers will catch on and we will be able to buy more packs from O/S.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

That gives me a great idea for taking batteries to the UK - I will put them into zippy boxes with a switch & LED and call them flashlights (or torches for the UK customs people )
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

Actually thats exactly what it would be. Lipo in equipment is different to single battery.

Postage restriction loop hole ftw.

Steve
_________________
Steven Martin
Twisted Constructions
http://www.botbitz.com

more wheel testing:

The flex in the printed tires is uneven due to using the slicer's linear in-fill pattern. To improve on that I drew up some radial patterns to test:



I have high hopes for the 'nautilus shell' spiral pattern as it has been used in many commercial robots, meaning the design is proven and just needs tweaking to suit Mr Mangle. Even though the tires are the same size as before, the printing time has shot up to almost 31 hours . No instant gratification for wheel testing this week.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

What size wheels are you after Nick, you might be able to find some squishy/conforming ones off the shelf.

The wheels are 90mm diameter by 32mm wide. I had a very careful look at all the options at HK and some other stores and there was just one tire the right diameter. It may be possible to use smaller wheels with some changes to the frame to increase the ground clearance.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

One of the options for commercial wheels is to find one with a really good tread pattern and grip and then cut out just the face of the tire and fit it over a printed core. That would give me a wheel with plenty of grip, a replaceable outer surface and the suspension that Mr Mangle needs to drive straight - too much to ask for?

There are dozens, maybe hundreds of good wheels with a diameter between 80 and 84mm, so if I can alter Mangle to have an extra 5mm ground clearance at the rear, then this idea could just work.

Can anyone suggest a good scientific way to test the grip of different tires on an arena floor? I am thinking of getting a selection of tires with different rubber compounds and tread patterns to find what actually works best.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

You could make a rig that uses a slow high torque gearbox connected to your drive wheel to take a weight up a 30 degree incline made of floor material. Measure how much weight you can load on each tread design before it slips so much it takes longer than 10s to climb the incline. Do, say, 30 trials of this for each tread design and check for statistical significance with ANOVA. Or something of that nature.

Wow, that's a bit more scientific than I was thinking . Statistics is my Kryptonite; anything more than an average causes a nasty rash and plotting results on a bell curve will probably cause organ failure.

The hill climb idea is pretty good; its a real-world and repeatable test. Time to place a wheel order.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

I know, stats often hurts, but you mentioned science and unless you do multiple trials for each tread and understand how they vary with respect to each other you may as well choose the tyres based on the colour of the rubber (my vote would be for yellow if you can find it ).

If you don't want to do fancy things with numbers, then you could always just plot all the results on a piece of graph paper and just check that the results of the winning tread design don't overlap the results of any of the other tread design by much. The confidence interval for an experiment like this is usually taken to be 95%, which means that if you do 20 trials of each tread, you shouldn't expect any more than 1 of the worst results from the best tread to be worse than the best result from the any of the other treads in a statistically significant result (of course you should ideally do a lot more than 20 trials).

For a friction test the traditional method is use a horizontal surface.
Put the test sample with a set weight on top of it on the surface with a string running over a pulley to a stack of weights hanging off the bench.

Static friction can be found by slowly adding weight until the test starts moving.
Dynamic friction can be found by adding weight then nudging the test sample.
*generally* for rubber on stuff dynamic is ~10% less than static.

In theory the tread design should make no difference to the coefficient of friction, as you reduce the surface area you increase the surface pressure. In practise there will be differences but they aren't going to be massive. For automotive tyres they are concerned with things like water removal and road noise but most importantly the marketing dept needs to sign off on the tread pattern.
_________________
Mechanical engineers build weapons, civil engineers build targets

More good ideas! Here is the simplest test rig I can think of, using parts on hand. Rather than moving the wheel, I will pull a strip of floor material past a fixed wheel:



The scales on the floor strip will show the force in Kg needed to pull the strip and while it may not be super accurate, it should be repeatable and show the relative grip of the wheels. Averaging out a number of tests for each wheel will improve the results.

I ordered four wheels with different tread patterns and rubber softness; the softest, smoothest wheels will probably be the winners unless the rubber wears out too fast. Most of the wheels are 84mm diameter and on is 90mm (the ideal size) but its the cheapest & lowest quality one.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion

The smaller wheels will cause the back of the frame to drag on the floor unless changes are made. The original design had the top tapering down at the back to save weight; now the top is flat the the bottom of the frame is tapered:



It looks very strange to have a bend in the base of the frame but all the parts fit, the weight stays virtually the same and there is plenty of floor clearance. The base plate will be made in two parts and welded across the frame - more work but not a big deal.
_________________
Australian 2015 Featherweight champion
UK 2016 Gladiator champion