Sumo robot challenges are fun. One important aspect of every Sumo robot are the tires: if they are sticky enough, the robot can push out the opponent. In this article I compare different available robot hubs and tires, and how to make DIY hubs and tires.
Outline
Sumo robots are great for teaching robotics. Our very special robot based on Pololu chassis is used at the Lucerne University of Applied Sciences and Arts (see “Zumo Robot with WiFi and GPS“) in education and robotics research.
While this robot is great, the used components makes it a rather expensive robot in the range of $150-180 (depending on options).
For STEM activities and conferences we are looking for a less expensive and stripped down variant, ideally in the $50-70 range (depending on options and parts used). The robot shall be cost effective, but still competitive in Sumo robot battles.
The idea is to use lower cost components, and a first concept of such a robot has been presented in “New Concept for 2018 Mini Sumo Roboter“:
Because the Pololu Zumo chassis and wheels accounts for $20 in the current design, one idea is to use 3D printed hubs and custom tires. And this is what this article is about: wheels and tires 🙂
Commercial Sumo Wheels and Tires
I looked at different Sumo wheels and tires available on the market.
Solabotics RW2
The first one is the Solarbotics RW2 Wheel (Item #642) from Pololu. I ordered the #642, but the #1127 wold be better as it has an internal set screw. But not a problem for me, as I 3D printed the hub. The wheels are sold individually ($3.95), so a pair is for $7.90.
For the Solarbotics I’m using a custom 3D printed hub:
JSumo SLT20
JSumo SLT20 Silicon Wheels costs $16 as a pair with aluminium hubs (they have a steel hub version for $18).
For the JSumo tire I have created a 3D printable hub:
FingerTech Cobra
FingerTech Robotics sells pairs of the ‘Cobra Minisumo Wheels‘ for $17.37:
Pololu Zumo
In comparison, this is the current robot we use with the Pololu tracks: The chassis with the wheel costs $19.95 (Pololu #1418). The picture below shows our robot with custom microcontroller board:
DIY Sumo Tires with Polyurethane
I did some research and found out that it is possible to cast custom tires using polyurethane. I’m using a Shore 20 material from https://www.polytek.com/ with optional C part (softener). One A part is mixed with 2 parts of B (by weight). Around 10 cl are used for single tire, so the 3 litres are good for up to about 300 tires:
To mold the tires, I have created 3D printed models:
Preparing the first mold:
mixing only a small amount and filled in the tire mold:
I let it cure for 16 hours. But the first one did not came out great:
First lesson learned: Part A and Part B were not mixed good enough, the material did not cure good enough :-(.
Second lesson learned: The rubber sticks too much to the mold as I did not use any release agent 😦
For the second try I mixed the components better.
Again, letting it cure for 16 hours, and the result looks better:
Additionally, I used Glycerine as release agent, and that worked very well:
The DIY tires came out nearly perfect: They are pretty soft, so I think I don’t need that C softener part at all.
Here with the tires attached on the robot. The same hubs as for the JSumo wheels can be used:
What is better? Below my impression of stickiness of the different tires, with 5 points the most:
- FingerTech: 5 points, very sticky
- JSumo: 4.5 points, very sticky, a bit less than the FingerTech one
- DIY polyurethane: 4.5 points, about the same as the JSumo one, softer than the JSumo
- Solarbotics: 4 points, not as sticky, but still good
- Pololu Zumo: 3.5 points, the tracks don’t have much stickiness
To measure the push force, I used a digital scale the robot has to push against:
Both robots (2-wheeled and Pololu Zumo tracked robot) used the same 1:75 DC motors. Sumo robots can be up to 500 g in weight.
But at that weight the robot with the DIY, JSumo and FingerTech tires did block the motors. To have the most traction I had to reduce the robot weight to 250g. Below the results (average over 5 runs):
- FingerTech (250g): 820g
- DIY Polytek A20 (250g): 810g
- JSumo (250g): 780g
- Solarbotics (500g): 750g
- Zumo (500g): 710g
In some sense, it is just a rough comparison. With higher weight the FingerTech/DIY Polytek and JSumo tires could push much more, but would need a higher DC motor gear ratio.
Costs
The stickiness of FingerTech/JSumo/DIY Polyurethane tires are about the same. Not counting the costs of 3D printing, the DIY Polyurethane is the least expensive one.
Below a summary for a pair of wheels with the least expensive shipping option:
- FingerTech: $17.37 (+Shipping $10.00)
- JSumo SLT20: $16.00 (+Shipping $40.95)
- Pololu Solarbotics: $7.90 (+Shipping $18.95)
- Pololu Zumo Chassis: $19.95 (+Shipping $18.95)
- DIY Polytek polyurethane (3 Liter for 150 tires): $45 (+Shipping $15)
I needed 10 ml Polyurethane for a tire Tire, so I could make up to 300 tires or tires for 150 robots. Ordering 50 or more of above hubs and wheel sets would give about 15% rebate and shipping costs would be divided more. The DIY Polyurethane tire molds would be the least expensive option, but does not factor in labor time and the 3D printing of the hubs. Otherwise the JSumo SLT20 is a good option: careless, very good hub and tires, but still rather expensive.
Summary
Happy Tiring 🙂
Links
- Pololu: https://www.pololu.com/
- JSumo: http://www.jsumo.com/
- FingerTech: https://www.fingertechrobotics.com
- Polytek: https://www.polytek.com/
Pingback: Custom 3D Printed Magnetic Encoder Disks for Robotics Projects | MCU on Eclipse
hello
could you share the tire mold drwings?
LikeLike
They are available with the rest of the project: https://github.com/ErichStyger/McuMiniSumoRobot
LikeLike
How much push force can the tyres of your bot resist? (Like the force till which the bot does not move)
LikeLike
that mostly depends on the weight of the robot itself.
LikeLike
Hi, for a conventional sumo(3kg) what shore A number would you recommend? 40-50A perhaps? Thanks for the guide!
LikeLiked by 1 person
Have not used that class. The 40 range sounds good to me, but you really have to test it. And it will depend on how thick your tires are.
LikeLike