The Axle Dilemma: A study in Fixed vs Loose axles

There are many decisions that go into modifying a diecast race car.  Each one of those decisions has an impact on your car's performance.  I am relatively new to 1:64 racing and still have many questions and a lot to learn.  There are a number of resources online with countless opinions and ideas about what makes a diecast car fast.  However, I have found that most of these opinions are exactly that...opinions.  There are some great videos about the more established world of Pinewood Derby racing, which cite evidence for why certain adjustments make your car faster, but not all of the science is translatable to racing diecast cars.  Primarily, what I like to call, "The axel dilemma."

It seems the most widely held opinion is that securing your axels (fixed down with JB weld) will make your car faster than leaving them loose.  However, I have heard of a few experienced and successfull builders who prefer to run their cars with loose axles.  So which is better?  Why?  I decided to do an experiment to see for myself.

I bought four current mainline Hot Wheels.  Two Ford Focus RS and two 2017 Acura NSX.  I labeled each set of cars 1 and 2 in order to tell them apart.  Initially, right out of the package, I ran each set of cars down my modest track at home.  20 rounds, alternating lanes.  The results were almost identical.  Ford1: 11 wins, Ford2: 9 wins.  Acura1: 9 wins, Acura2: 11 wins.  The stock cars were very similar in speed and performance.

My test track starts with an Orange Track drag section, then turns into Crash Racers open track with a 180 and a loop.  The inside lane is adventageous, so if the cars are equal, whichever one starts there usually wins.  A car can, and does win from the outside lane if it is fast enough to overtake the inside car and beat it to the first corner.

After the initial test race, I drilled out all four cars.  I fixed the axels of Ford2 and Acura1 with JB Weld.  I left the axels of Ford1 and Acura2 loose.  I purposely chose the Ford2 and Acura1 beacue their results were slightly worse than Ford1 and Acura2.  I guessed that if fixed axels are better, the new advantage would be easliy seen in the "slower" cars.  I left out the interiors of the fixed axels cars, but made no other modifications.  Then I secured all four cars back together with screws.

I re-ran the cars, 20 rounds like before, alternating lanes.  The results were staggering.  Ford1: 19 wins, Ford 2: 1 win.  Acura1: 6 wins, Acura2: 14 wins.  There was a marked decrease in speed in the fixed axel cars.  They consistently did not perform as well in the inital drag section of the track and did not beat the loose axel cars to the first corner.

Most surprising to me, was how much slower Ford2's fixed axel wheels seemed to spin.  Initially, when I eyeball tested the cars, spinning each wheel and seeing how long it would turn, they all seemed pretty much the same.  After securing Ford2's axels, the wheels turned for a noticably shorter time when I spun them with my finger.

This is a very small sample size, and by no means comprehensive, but I was surprised at the results.  Has anyone else done experiments like this?  What has been your experience?  

I plan on doing more experiments like this.  I still have lots of questions: What actually makes a "good wheel", is silent really better?  (I've found some really fast, noisy cars)  Does sanding wheels make a significant difference?  (I hope not, because I find it really tedious)