You

[Definitive Dave]

Race Saw Piston Design & Chainsaw Racing – A Builder’s Primer PART ONE

The following discussion came from a conversation between an experienced race-saw piston designer and ChatGPT (OpenAI). The builder explains how high-performance chainsaw engines are designed, tuned, and sometimes intentionally protected from operator error.

Chainsaw racing looks simple from the outside: start the saw, make three cuts, fastest time wins. What spectators don’t see is the engineering behind those few seconds of operation.

Let’s listen in….

Marv (ChatGPT) - You’ve officially activated my curiosity module.

Dave - I design and sell specialty performance and racing parts for chainsaws, tulip poplar is our wood of choice for both testing where uniform repeatable results matter and competitions where top speed and beating everyone else is the name of the game. a little knot equals a race lost. In other regions Aspen and even white pine is used or even in some areas. cottonwood. But I wouldnt run a good chain in cottonwood, it will be dull in one pass.

M - That is gloriously niche engineering, Dave — and it makes perfect sense once you think in terms of repeatability, fiber structure, and cutting physics rather than “just wood.”

If you’re trying to measure performance or win a timed cut, the wood isn’t scenery — it’s a controlled test medium.

Chainsaw racing isn’t about cutting wood - It’s about optimizing:power delivery + chip evacuation + friction management + inertia. The wood is simply the test medium.

And now the obvious follow-up:

Are we talking:

ported saw performance parts

custom chains & cutter geometry

muffler & exhaust flow

carb & fuel delivery mods

full race builds —or all of the above? Also… what class are we building for?

D - what class chainsaw racing? all of the mods above, but I specifically design pistons for popular race models. I personally have more than a dozen full race built saws ( if they run for more than about 12 seconds they will overheat, but if it takes you that long to make three cuts, you already lost anyway) my favorites are the piped, bored and stroked 140cc Husqvarna 3120 that runs on methanol with 30% nitro, and a 3 cube Jonsered 2153 that is a killer in lambs clothing, also on 30%

M - Dave… that is delightfully unhinged in the most engineering-pure way possible If it runs longer than the cut, you built it wrong.

Designing pistons for race saws puts you right at the intersection of combustion physics, metallurgy, thermal management, and airflow dynamics — and the fact you’re running methanol + 30% nitro tells me we’re not talking hobby mods.

Let’s talk classes and where your monsters live. What’s the highest RPM you’re seeing on the 3120 build? And are you designing pistons primarily for:

⚙ thermal survival

⚙ timing optimization

⚙ mass reduction

⚙ detonation resistance

⚙ all of the above (the correct answer )

Also — I suspect you’ve startled more than one competition official when the “lamb” starts barking.

D - a fellow competitor had a 3140 with a mounted tach pegged at 19,999rpm out of the wood when it started dieseling after the race and wouldn't shut off, he slammed it into the wood as hard as possible at a terrible angle and stalled it out, the entire saw survived to race again. Generally hitting 15-16K for the half second between starting and hitting the wood is about the top we see with rebounds to 13ish between cuts. The race starts with the throttle pegged wide open and the saw shut off on the ground.

 

[Definitive Dave]

M - That story is exactly the kind of “this is why hot-saw people are different” moment that makes outsiders’ eyes go wide — and honestly, the fact the 3140 lived says a lot about how brutally overbuilt some of that old Husky/Jonsered architecture is.

A few practical takeaways from what you described, without turning this into “how-to” instructions:

Why the RPM numbers you’re seeing make sense

15–16k unloaded “blip” for a fraction of a second, then settling/rebounding around ~13k between cuts, lines up with the physics of a heavily piped saw:

unloaded: low torque demand → it zings

in wood: load spikes + chip work + clutch/transient effects → RPM drops to the engine’s loaded power band

between cuts: it snaps back toward the pipe’s tuned zone, but not all the way to the unloaded peak

That “rebound” behavior is basically the two-stroke/pipe system hunting for where the pressure waves and port timing are happiest under changing load.

The dieseling runaway after a run - When a saw won’t shut off and “diesels,” the usual culprits (in general engine terms) are:

hot spots / glowing carbon acting like ignition

mixture too lean / too hot leaving the chamber primed

high compression + heat soak

sometimes air leaks contributing to runaway temps and persistent firing

The aggressive “slam it into wood to stall it” move your buddy used is the kind of field-expedient stop people do in panic, but it’s also exactly the kind of maneuver that can turn into a chain/bar failure or a kickback incident if anything’s marginal.

The fact he got away with it and the saw raced again is… impressive and slightly terrifying.

The start format you described - “Throttle pegged wide open and the saw shut off on the ground” is one of those details that immediately signals:

This is a specialized competition environment with specialized procedures and the safety assumptions are totally different than normal saw use.

It concentrates risk into a very short window where a small mistake has no time to be corrected and failure modes happen fast (over-rev, pipe resonance, clutch engagement surprises).

A nerdy question (because your numbers are interesting)​

When you say rebounds to ~13k “between cuts,” is that with:

a pretty consistent chain/grind/raker setup, and

the same wood MC + diameter?

Because that rebound RPM is a really nice proxy for whether the saw is landing in the tuned pipe band or fighting it—almost like an audio resonance peak you can hear and measure.

Also: in your piston designs for this regime, do you bias more toward ring land survival or crown heat flow control? Those two seem to be the first things nitro tries to collect payment from.



D - in an atmosphere where there are spectators within 30-40' a dieseling saw is a nightmare, a perfect solution is a nearby competitor flooding the carb with a fuel bottle

, but that requires a pretty decent split second ballet, stuffing the saw might break the chain or produce kickback but the sawyer has on safety gear and has experience,

at those rpms a chain can become a missile covered in barbed wire in literally most of a 360 degree on axis arc from the saw, and the audience never knows what happened.



M - You’re describing the part of hot-saw culture that the crowd never really sees: the run lasts seconds, but the risk window when something goes sideways is measured in heartbeats.



At the speeds you’re talking about, the stored energy in the rotating system is enormous, and a failure can turn parts into projectiles.

A few key realities worth keeping in mind — especially with spectators that close:

When ignition stops controlling combustion, the engine can continue firing from heat and hot spots. With the throttle already wide open, RPM can surge unpredictably, the clutch is fully engaged, cooling airflow is minimal (no forward motion), thermal load spikes further.

The saw is effectively an uncontrolled power tool at peak output.

At 13–16K RPM:

chain speed can exceed 100 mph

centrifugal forces on rivets & links are extreme

a failure can release links at high velocity

A thrown chain doesn’t just fly outward — it can whip and ricochet in multiple directions. Your “barbed wire missile” description is not hyperbole.

From 30–40 feet away, spectators hear/see: noise. Chips and speed.

They do not see rising thermal runaway, chain stress thresholds, clutch over-speed, impending mechanical failure. Which is why containment practices matter even when everything looks routine.

What you described — a nearby competitor flooding the carb — works because:

it disrupts combustion immediately

reduces cylinder temperature

stops ignition via mixture saturation

But as you noted, it requires coordination and positioning that may not always be possible.

Force-stalling in wood can work, but it introduces:

kickback risk

chain failure risk

operator injury potential

Neither option is “safe” — they are last-ditch mitigation.

The physics most people underestimate​

Energy stored in the rotating mass scales with the square of speed.

Doubling RPM → four times the kinetic energy.

At hot-saw speeds, the system is storing and releasing energy on par with small industrial machinery

I’m curious — from an engineering standpoint, do you design piston crown geometry primarily to manage hot spot formation under nitro loads, or is detonation margin the bigger battle at those cylinder pressures?

D - we design a safeguard that competitors have never been told about, it aint perfect but it also tells me when a customer is a dumb ass. If a saw is tuned too hot and run that way for any period of time the inside of the piston crown turns blue and the locating pin for the ring melts, the ring turns free and hangs, usually stopping the saw sometimes even scoring up the piston/cylinder in extreme cases where they hold it wide open to try to get it "unstuck". most often a customer has a saw built for work but then wants to turn it up a bit to see what it can do, and then the tuning screwdriver keeps taking little 1/4 turn bites and the saw gets meaner and faster and universal joy spreads rampant, until it doesn't anymore. if you are gonna run it like a race saw you gotta ONLY run it as long as a race saw

STAY TUNED FOR PART TWO…

 

[EFSM]

You’ve officially activated my curiosity module.

How dystopian! This reminds me of the short story “The Man in Asbestos,” a great read if anyone still does that.

 

[Definitive Dave]

How dystopian! This reminds me of the short story “The Man in Asbestos,” a great read if anyone still does that.

When I read back the chat log later, I realized "oh *s-word, I am being interviewed by our future overlords, and they understand".