During the off-season I wanted to expand our capabilities by fabricating a custom gearbox. For simplicity I am just making a two CIM gearbox to start. I figured out the ratio I need and found all the gears off of andymark. The problem I have is figuring out far apart to space the gears for thermal expansion. And also do I use the gear pitch diameter to space them or is there some other measurement I need. None of our mentors have experience with this and I was wondering if any of you guys could help.
Thank you
You will want to use the pitch diameter of the gears to space them.
In terms of thermal expansion, I can’t help you there. You can take a look at the equations for thermal expansion and the materials using and see if the expansion is on the order of magnitude of your machining precision. If the thermal expansion you’re expecting is +0.0005" on the OD, but your machining tolerance is ±0.005", then it might not matter so much.
The rule of thumb is add .003 to center to center for FRC designs; there are caveats…
Use the pitch diameter of the gears. Add the two pitch diameter up, divide by 2 and then add your center to center modification.
Example: 20T and 80T 20dp gears
(1in + 4in)/2 + .003 = 2.503in center to center distance
Here is a little more clarity than just using an online calculator. It is really easy to calculate the center to center distances for gears
Here is the formula for center to center distance.
("#of teeth of gear 1"+"# of teeth of gear 2") / (“pitch Diameter” * 2) + “gear mesh spacing tolerance” = "center to center distance in inches’
For our team we use .003 for gear mesh spacing on 20dp gears and .0015 for 32dp gears.
So for example of a 20dp gear spacing of a 20 tooth and 60 tooth gears would be
(20+60)/(20*2) + .003 = 2.003 inches
An example for the calculations of a 32dp gear spacing for a 12 tooth pinion and a 60 tooth gear would be
(12+60)/(32*2) + .0015 = 1.1265 inches
Hopefully this helps
Look at coefficients of thermal expansion here: https://www.engineeringtoolbox.com/linear-expansion-coefficients-d_95.html
First, if your gearbox and gears are the same material (i.e. aluminum) then CTE won’t matter. As the box changes the gears will change the same amount.
If they’re different materials, or one is heating up much more than the the other, then let’s do a sample calc. I’ll assume that your gearbox is aluminum and you’re using the gears in tim-tim’s example, ctc spacing of 2.503, and the gears are steel with an aluminum gearbox:
Al - 22ppm/°C
Steel - 12ppm/°C
To see a 0.001in change in ctc spacing, which is about as accurately as many teams can make gearboxes in the first place:
2.503in10ppm/°Cdel_T=0.001in
del_T = 0.001in/2.503in/10ppm/°C = ~40°C. That’s the difference between room temperature and ‘so hot if you touch the assembly you’ll do tissue damage in moments.’ In other words, for FRC, you’ve got bigger problems! I would not worry about CTE.
I agree with your overall conclusion, that CTE in the FRC scope doesn’t really matter, but this statement is assuming the gears and plates are all the same temperature, which in all likelihood will not be the case.
For whatever it is worth, years ago I contacted Boston Gear regarding this question. I consider them an expert in this matter. Their answer was to use the sum of the diametral pitch radii to properly space the gears. No need to add more distance between the gears. Do what you want, but I space them this way without problem. When spacing sprockets, I again space the sprockets according to calculations, and ALWAYS provide a method for additional tensioning as needed. :] :]
This is correct, except gear profiles were designed to allow for a little wear in over the life of the gears, and get a little more efficient with a little run-in. FRC robots aren’t used long enough for this designed run-in to occur, so you can achieve a similar effect by moving the gears a little further apart. This way you can start with a slightly more efficient (quieter / smoother) geartrain at the sacrifice of additional backlash.
The other reason to add a little extra space is machining tolerances. A lot of teams struggle to hold tolerances within 1-2 thou unless they are using a CNC. And even with a CNC machining the plates, it’s easy to get plate misalignment that’s more than that (e.g. two plates that are clocked a little bit because they don’t have good alignment) unless you’re explicitly designing in tight-tolerance alignment features. Designing in a little extra space is an easy way to ensure your gear teeth don’t bottom out and jam.
Sum of the diametral pitch radii?
Could you go into more detail or provide an example?
Pitch diameter for a given gear is (# of teeth)/(diametrical pitch). Diametrical pitch radius is half of that (radius vs. diameter), or (1/2)*(# of teeth)/(diametrical pitch)
So the exact C-C distance for a given gear pair would be d=t1/(2dp1)+t2/(2dp2). Since the diametrical pitch of a gear pair must be the same for the two gears, you get the formula d=(t1+t2)/(2*dp)
For example, if you had a gear pair consisting of a 20t and a 50t gear (20dp), the exact C-C distance would be (20+50)/(2*20)=1.75". It’s standard in FRC to add 0.003" to that for reasons listed above, so your final distance would be 1.753".
This was explained in a separate post above and a calculation that I’ve used extensively.
The phrasing “sum of diametral pitch radii” was a bit confusing. I can understand where Teched3 was coming from when he contacted Boston Gear and assume that his “0.003 spacing” was not something they recommended, only the exact c-c.
I’ve honestly never heard of the radius of the pitch diameter referred to as the diametral pitch radius. I’ll add that one in my jargon notebook.