What Is Thermal Expansion?
Steel grows when it gets hot. In stamping die work, that growth causes problems.
I've built dies for 18 years. The shops I've worked in have lost thousands of dollars to parts that came out wrong because somebody didn't account for heat. A progressive die running at 600 strokes per minute generates friction. That friction raises the temperature of the punch and die steel. The steel expands. The clearances you ground yesterday are no longer the clearances you have today.
The Numbers
Tool steel expands at roughly 11 to 12.5 micrometers per meter for each degree Celsius of temperature change. That sounds small. On a 200mm die section, a 30°C temperature rise adds about 0.07mm to your dimensions. You just blew your tolerance on a part that needed ±0.05mm.
D2 steel, A2 steel, and most air-hardening tool steels fall in that 11-12.5 range. Carbide is lower, around 5-6. The base metal you're stamping has its own coefficient. Aluminum expands nearly twice as much as steel. Stainless sits somewhere in between.
Where It Shows Up
Wire EDM is the first place most die shops see thermal expansion cause real trouble. The dielectric fluid controls temperature, but the workpiece still heats up during cutting. A block that measured 150.000mm in the morning might read 150.015mm by afternoon. Some shops won't take final measurements until the part has sat overnight.
Grinding is worse. The wheel generates heat at the point of contact. Without flood coolant, a surface grinder can put enough heat into a die insert to throw off a bore location by several tenths. Jig grinding operations use temperature-controlled coolant systems for this reason. The better shops keep their grinding fluid within 1°C.
During production runs, the die heats up in a predictable pattern. The first 50 parts often show different dimensions than parts 500 through 1000. Experienced press operators know this. They'll scrap the first shots or adjust the press shut height after warmup.
Die Assembly Problems
A die set built at room temperature behaves differently at operating temperature. The punch holder and die shoe expand. The guide pins and bushings expand. If the materials don't match, the fit changes.
I've seen shops use bronze guide bushings in steel shoes without thinking about the thermal mismatch. Bronze expands more than steel. Under heat, the bushing gets looser in the bore. The die starts walking.
Dowel pins present a similar problem. A slip-fit dowel at 20°C becomes a loose dowel at 40°C. Some shops use a tighter interference fit to compensate. Others accept the movement and control it with heel blocks.
What Works
Temperature-controlled shops help. Keeping the ambient at 20°C ±1°C makes measurement consistent. Most tool rooms I've worked in were not temperature controlled. You learn to work around it.
Letting parts stabilize before final machining matters more than most people admit. Rough a part, let it sit 24 hours, then finish it. The part relieves internal stress and reaches equilibrium with the room. Your final dimensions hold.
Consistent startup procedures on the press reduce part variation. Run the die for 15 minutes before keeping parts. The steel reaches operating temperature. The clearances settle.
For dies that run hot-high-speed applications or heavy-gauge material-cooling channels drilled through the die sections pull heat out. The automotive industry uses these extensively for hot stamping operations where the blank comes in at 900°C. Water or oil circulates through passages machined into the die faces.
Compensation
Some die designers add compensation to their prints. They cut the wire EDM profile 0.01mm undersize knowing the part will grow in service. This approach requires production data. You need to know exactly how hot the die gets and how much growth that causes.
On progressive dies with multiple stations, each station may expand differently. The forming stations generate more heat than blanking stations. The strip layout has to account for cumulative effects.
Most shops don't calculate thermal compensation. They build the die to print, run test parts, then spot and adjust until the parts come out right. This works. It takes longer.
What Doesn't Work
Ignoring the problem doesn't work. I've watched die makers spend days chasing a dimension problem that turned out to be thermal drift in the grinding room. The AC unit cycled on and off. The room temperature swung 5°C. Every part came out different.
Trusting the CMM without checking room temperature doesn't work. Coordinate measuring machines compensate for their own thermal expansion. They don't compensate for the part sitting on the granite.
Rushing final inspection doesn't work. A part fresh off the wire machine is not the same size as a part that sat on the bench for four hours.
The Bottom Line
Thermal expansion is physics. Every die maker knows about it. Not every die maker plans for it. The shops that control their environment, stabilize their parts, and follow consistent procedures get better results. The shops that don't spend more time adjusting dies in the press.
The numbers are in the handbooks. The coefficient for your steel is a Google search away. The hard part is building habits that account for heat in every operation, every day. After 18 years, I still have to remind myself to let parts cool before I measure them.