Shot Peening Process Control: How to Stay in Spec in Industrial Manufacturing
Updated: Oct 7
Controlling the process means achieving the Almen intensity required, on the strips chosen for the parts in production. Here below a case study first, and the general concept follows.
Case study: Almen control on a tray-loaded peening line
The parts ran on trays. That was a handling decision, not a peening one — trays made loading, moving and unloading manageable, and the machine was fed one tray at a time, parts on top, processed, unloaded, reloaded.
A few of those trays were modified to carry Almen strips.
That is the whole intervention. A control tray loads like any other tray, runs like any other tray, unloads like any other tray. Production never stops, nothing is set aside, no one waits for the machine. The test happens inside the normal flow, not beside it. Between production cycles the tray is prepared with the Almen strips, and after the run they are measured and reported.
The site runs 24/7, so the control tray goes through twice in twenty-four hours. Those two readings sit alongside the checks the machine already gives you — wheel amperage above all — and together they cover the two things that can move: what the machine is drawing, and what the process is actually delivering on the part.
Five years of production under this control. Not once did a part leave the plant incorrectly peened.
What makes a peening control industrialisable in manufacturing
Context decides everything; trials are done in laboratories, by skilled technicians and engineers with all the background they need. Real production life is totally different. If you have ever worked on a production floor, this needs no further explanation. So here is what to check before making it, and why.
Machine requirements for peening process
The machine must have good amperage control on the wheels, and the related limits control in it. This is not optional: wheel amperage is how you measure the energy the wheels are actually delivering, through the media, onto the parts and onto the strips.
Then it depends on the machine's layout:
Batch machine — a reliable timer on the cycle
Pass-thru machine — equally reliable control of the pass speed
None of this is science fiction. Machines have had these controls for twenty years at least, and where they are missing, any industrial electrician can fit them for a cost that is close to ridiculous. Do not let the machine requirements put you off: a sound machine can do the job perfectly well in industrial peening.
Production team requirements to apply shot peening in manufacturing
Be realistic about your production team, and start from turnover. If you lose two people a year out of fifteen, the ones who stay will train the two you hire. If you lose ten every couple of months, keeping a minimum level of knowledge inside the team becomes a problem in itself — and it is a problem the check has to be designed around, not one you can train your way out of.
Nobody can make that assessment except whoever runs the production process as a whole. It is not a technical judgement, it is a judgement about your own floor — and it has to be an objective one. The usual answers are that the team is excellent, or that the team is hopeless. In my experience most floors sit somewhere in the middle, and a small amount of work moves them a long way. That is why realism matters more than optimism here: it is the only way to see what is actually worth doing.
The knowledge involved is not extraordinary, far from it. But it has to be passed to the production team, and you have to guarantee — to them and to yourself — that it stays effective over time.
A simple example. If the strips are read badly, you can spend a week trying to work out what is wrong with the process, when nothing was wrong with the process at all. It is a handful of operations. They have to be done the right way, or the whole thing becomes very hard to manage.
Process requirements
Once you know what the machine can do and that the minimum requirements are met, and once you have an honest picture of the team — ordinary operators, nobody exceptional, but people who would rather do a job properly and safely — you can design the check itself.
It has to be effective, efficient, cheap, and it has to be runnable by any member of the team without significant variation in the reading. Same number on any shift, at any hour, by any operator. Which is the definition of quality.
A separate note, and a strong one. The best protection against peening scrap is to put the control into the machine itself, in requirement one: a program that stops the process when amperage drops below threshold. No judgement call, no shift to wait for. The machine refuses to run out of spec. To all effects this is jidoka applied to the peening machine and to the peening process — the machine detects the abnormal condition and halts on its own.
Add one thing to it. If the process was thought through properly, with the right media, over-peening is generally not a problem: a part can take a doubled cycle without harm. Which means a part caught by the stop can simply be run again, and scrap on peening goes to virtually zero.
How the procedure gets built in the first place — from the part to the media to the strip positions — is in the article The Shot Peener published on this.
What if the peening process is detected going wrong?
A process, obviously. You need to build the appropriate procedure to identify what is really going wrong and indeed if it is really going wrong. Read carefully the following, it will save so many headaches and weekends that you can't even imagine.
The first move is not to look at the machine. It is to question the measurement. Faulty Almen strips exist, but far more often the test was simply run badly. So: take the person who ran it off the job, and have two different people run it again. If the second round comes back in spec, the first one was wrong. Run it a third time to confirm, and you are done.
If the second round confirms the problem, production stops.
Then the plan says who gets called to look at the machine, and that person works to one criterion only: this machine, with the media that is in it, has to produce a given Almen reading. Which means it has to throw a given number of kilos per minute over a given time — which means a given amperage. That is what they work on. Nothing else.
Typically the problem is either in the machine or in the parts. It does not go much further than that.
The hours lost chasing a faulty strip or a badly taken reading are beyond counting.
Peening data processing
Three strips, four times a day, is twelve numbers. Eighty-four a week. Three hundred and thirty-six a month. Over a year you are handling something close to four thousand readings, and that is on a modest control plan.
The number itself comes off the gauge — a dial indicator reading the arc, in thousandths of an inch or hundredths of a millimetre. One number at a time, taken by whoever is on shift.
What matters is not the single reading. It is how those numbers move over time. A process rarely fails in one step; it drifts, and the drift is visible in the series long before any single reading breaks a limit. That is also what tells you when the machine needs maintenance — and lets you plan the stop instead of taking it.
So the handling of these numbers has to be designed, from the moment the reading is taken to the moment it is archived, by whoever is responsible for the process or the production. In some cases the data has to be kept alongside the production logs, because it is what gives the batch its traceability.
In theory a spreadsheet does all of this. In practice spreadsheets have a long history of quiet damage in manufacturing — a copied formula, a column sorted without the others, a file that lives on one laptop. They work, but only with discipline behind them.
You can write a small program. You can do it by hand with real diligence. Or you can use Vortex, which is built to manage wheel blast and peening machines and has a shot peening section inside it: reminders so the operator knows when the test is due, data recorded at the point it is taken, uploaded to a server, and waiting in order for whoever reviews it — who can then export to Excel and work the way they prefer.
Conclusion: what Almen strips are not, and cannot do
Across this series the point has been that the Almen strip is an indirect instrument. It measures peening intensity, and it does it well enough to run a process on. What it does not measure is the residual stress in the part.
That limit is worth naming, because today there are ways to measure the real thing: optical examination under magnification, and X-ray diffraction for residual stress, in megapascals, on the actual component.
It is worth scheduling that verification at regular intervals, ideally externally, to certify that the whole system is still inside the parameters it was built on. It costs money. In some cases it makes sense, in others it does not — but in industrial manufacturing, with real volumes, the cost is negligible against the number of parts produced.
Think of the chain it closes:
Wheel amperage tells me I am delivering the same energy, through the media, onto the part and onto the strip
The arc on the strip is, by definition, the intensity the part is receiving
A periodic residual stress measurement on the part confirms that the whole chain still means what it is supposed to mean
That last step closes the loop. It lets you confirm with 2026 instruments a method engineered in the 1930s and 1940s — and that method is still the one running the floor. There is something remarkable in this piece of pure mechanical engineering.
Table of content:
4) Almen intensity, KPIs and how to industrialize the process. this post.
