If you have been around high-cycle machinery at all, you already know the cam follower roller is where a lot of machines quietly fail. It looks like an insignificant little wheel, but it carries the full cam load at several thousand revolutions per minute, day after day. I have opened machines where the follower was the only part that had actually failed, and the whole machine stopped because we skimped on a bearing that cost twenty euros. This article walks through the types, the load math, and the selection habits that keep a follower alive.
1. What a Cam Follower Actually Carries
A cam follower is a stud-mounted rolling element that rides against a cam profile or a track and transfers the cam motion to the follower arm. It is not just a bearing; it is the interface between the profile and the mechanism. It sees radial load from the cam, plus any thrust if the profile is not perfectly aligned, plus shock loads every time the profile changes dwell to rise. And it does this while a lubricant film is being squeezed and released thousands of times per minute.
Because the cam and follower contact has a tiny radius of curvature relative to the load, the contact stress at that point is brutal. That is the number that actually decides life, and it is why followers are made from case-hardened material with a precisely ground crown.
My blunt take: a follower fails from contact fatigue or lubrication failure long before it fails from pure load. Selection is about controlling contact stress and keeping the film alive.
There are two big families: the roller type with an inner race and needle rollers, and the stud type where the stud itself is the inner race. Within those you can order crown versus flat, sealed versus open, and with or without eccentric studs for adjustment. The catalogue pictures all look similar, but the differences decide the life.
2. Roller Type Versus Stud Type
The roller follower, sometimes called a needle roller track follower, has a through-bored inner ring that is mounted on a separate shaft or pin. The stud type integrates the shaft as a single piece with the rolling track ground directly into the stud. For most cam mechanisms, the stud type is the default because it mounts with a simple hex socket and locknut and needs no extra shaft hardware.
There is a trade you should know. The stud-type follower offers higher radial capacity for a given footprint because the section is optimized, but it concentrates the load into the stud bending and the mounting bore. The roller type spreads the load along a cleaner path but needs a correctly machined shaft and shoulders, which is harder to guarantee on a non-standard machine than on an OEM axis.
| Feature | Stud Type | Roller Type |
|---|---|---|
| Mounting | Hex socket + locknut | Shaft and retaining ring |
| Radial capacity | Higher per size | Good, mounting dependent |
| Camber/alignment | Less forgiving | More forgiving |
| Typical use | Cam, track, index | Large profile, OEM |
That alignment point deserves emphasis. A stud follower is rigidly tied to the bracket, so if the cam track is not square, the contact edge-loads and the crown cannot save you. I have seen a supposedly crowned follower die in a week because the mounting surface was off by 0.15 mm over the width, which put the whole load on one edge of the crown.
Check the squareness of the bore and the flatness of the bracket face before you blame the bearing. Honest advice: the bracket tolerance usually decides the follower life more than the catalogue life calculation does.
3. Crown Versus Flat Profile
The profile of the rolling surface, crown versus flat, is the selection detail people skip, and it is the one that protects you from edge loading. A crowned follower has a slightly convex surface, and the crown radius is a compromise. A small crown radius concentrates the contact but handles misalignment beautifully; a flatter surface spreads the contact and raises the load capacity but punishes misalignment.
Most cam followers are supplied crowned as standard, and catalogue values already assume some crown. The honest reading of the catalogue is: if your application has any real misalignment risk, keep the standard crown. Only go flat-track when you can guarantee squareness and want the maximum contact area for a predictable straight-line heavy load. In my experience, people go flat to gain a few percent of capacity and lose far more in edge-load sensitivity.
4. Loads, Speeds and the Life Calculation
Now the math, because the catalogue number you need is not the static rating but the basic dynamic load rating C, expressed at a defined speed. The life of a rolling element under line contact follows the standard power law, and for cylindrical followers the life exponent sits around 10/3. The practical formula looks like this:
L10 = (C / P)^(10/3) x 10^6 revolutions, where P is the equivalent dynamic load and C is the basic dynamic load rating from the catalogue.
The catch is P. The cam load is not constant; it varies through the profile, so the equivalent load must be calculated as a weighted average over the cycle. A common method is to break the lift curve into segments, compute the load and the time fraction for each segment at the contact point, and take the cube root of the sum of the cubed segment loads weighted by their time fractions. Doing this by hand is miserable, which is why every serious designer ends up with a spreadsheet.
Do not forget the speed correction. Catalogue life values assume a reference speed, and below it the life improves, above it the lubrication and heat start to hurt. A follower spinning at 6,000 rpm on a pick-and-place cam reaches temperatures that kill the grease; that is a reality you have to carry in, whether or not the catalogue says so.
5. Seals, Grease and the Lubrication Question
For the lubricant decision I have a simple rule based on where the machine lives and what the operator touches. Sealed and pre-lubricated followers with metal shields or contact seals last longest in dusty shops, but they reach their grease life limit and then it is a swap, not a relube. Open followers with a grease fitting can be relubricated on schedule, and on high-cycle machines a tiny weekly pump of grease is worth more than any capacity margin you bought.
Here is the trap that gets a lot of designers, including me early on: the follower must be relubricated with the grease specified for the speed and the temperature, not whatever tube is on the shelf. EP grease for sliding contact, a softer lithium soap for high speed, and no grease at all past the manufacturer speed limit, where relubrication intervals collapse and the whole thing becomes a wear test. The catalogue always has a relubrication interval table. Read it.
You can buy a follower that will never be relubricated and accept its sealed life, or a follower you can relube, but you cannot buy your way out of the temperature and speed reality.
6. The Mounting Details That Kill Followers
A follower is only as good as the bracket that carries it. Watch the bore diameter and its tolerance; the stud fits into a reamed bore with a light interference or a slight clearance depending on the catalogue guidance, and guessing wrong produces fretting or a spinning stud. Lock the nut with the specified torque, which is often in the 20 to 40 N*m range for common sizes, and always use the locking collar or washer that the catalogue actually calls for.
Make sure the spring load or the cam geometry keeps the follower in contact with the profile through the whole cycle. If the follower ever loses contact, it rejoins with a shock that takes a bite out of the race, and you will hear it as a ticking that no amount of grease fixes. I design the return spring so the worst-case deceleration never lifts the follower, and I check that at the fastest profile segment, not the average.
Parallelism between the cam track and the follower axis matters too. A few hundredths of a millimeter of skew shows up as a tapered wear band on the crown, and by the time you see it the race is already gone. That measurement belongs on the assembly drawing and the first-article inspection, because it will not fix itself.
7. A Worked Example From the Shop Floor
Let me walk through a real pick-and-place index I did. The choosing arm ran at 180 cycles per minute, and each cycle the follower saw a rise segment of roughly 45 degrees of cam rotation with a peak load I measured at 2.4 kN at the roller from the arm weight and the component. The first iteration used an 8 mm stud follower with a dynamic rating around 1.2 kN, because the original designer sized it on steady-state weight. On the third month the follower was scalloped and the machine sounded like a typewriter.
I replaced it with a 16 mm stud follower rated around 3.2 kN dynamic, checked the bracket squareness with a feeler over the mounting face, and added a grease nipple. The revised equivalent load P came out near 1.1 kN after the weighted cycle average, giving a calculated life of several years, and the tabular point here is how much of the gain came from the bracket fix, maybe a third of it, not the bigger bearing. People always remember the bigger bearing and forget the fitting and the square face.
| Parameter | Original | Revised |
|---|---|---|
| Follower size | 8 mm stud | 16 mm stud |
| Dynamic rating C | 1.2 kN | 3.2 kN |
| Equivalent load P | 2.0 kN | 1.1 kN |
| Bracket squareness | Not checked | Feelered, <0.03 mm |
| Calculated life | Weeks | Years |
8. Selection Checklist to Print and Keep
- Determine the peak radial force and the profile speed at that point.
- Select the follower family: stud for compact cam mounts, roller for large profiles.
- Choose crown standard unless you can guarantee squareness.
- Pick a dynamic rating with the calculated equivalent load and the target life.
- Specify the seal and grease for the environment.
- Check the bracket bore, torque, and parallelism, and put them on the drawing.
- Verify return force keeps contact through the fastest segment.
Keep the checklist taped to the drafting table and the failures drop. You will still have the odd unlucky follower, but you will stop having the systematic ones, the ones that come from hoping a catalogue number covers a bracket you never inspected.
There is one more thing I want to leave you with, and it is a habit rather than a calculation. Keep a failure log for the followers on your machines: the running hours, the grease used, the load at the time, and a photo of the failed race when you pull it. After a few entries the pattern jumps out, and you start predicting failures instead of chasing them. I have a folder of follower photos going back years, and when the vendor claims their bearing is better, I hand them the log. That exchange is worth more than any technical battle, because the vendor sees the actual contact fatigue patterns from your actual machines, and so do you.
