# The Furnace Chart That Saved a Production Line: Why Heat Treatment Paperwork Matters

> When a heat-treat furnace drifted 18°C, 4,000 A-286 bolts were quietly under-aged. The customer's incoming hardness check caught it, but only because they checked.


Last July, a customer's incoming inspection flagged a lot of A-286 bolts for low hardness. Our metallurgy team pulled the furnace chart. The chart showed the load thermocouple had read 702°C during the aging cycle, not the specified 720°C. The furnace controller setpoint was 720°C. The actual metal temperature, measured at the load, was 18°C lower.

The bolts passed tensile testing at room temperature. They would have shipped. They would have been installed. And they would have under-performed in creep at 650°C: not failing outright, but relaxing faster than designed and needing re-torque at every inspection interval for the life of the machine.

The customer's incoming hardness check caught the 18°C because they didn't rely on the MTC alone. Here's what they did differently.

## What "trust but verify" looks like for heat-treated bolts

Most customers accept the mill test certificate and move on. This customer's incoming QA procedure includes:

1. **Hardness test 3-5 bolts per lot.** For A-286 per AMS 5732, that's 24-35 HRC. Values near the bottom of the range are acceptable but warrant a closer look at the furnace chart.

2. **Read the furnace chart, not just the MTC.** The MTC says "aged at 720°C for 16 hours." The furnace chart shows whether the load thermocouple actually reached 720°C, for how long, and at what point in the cycle. An MTC that says 720°C is a statement. A furnace chart that shows 702°C is the truth.

3. **Cross-check the hardness against the tensile cert.** A lot with 24-26 HRC hardness that certifies at 1000 MPa tensile is physically possible but statistically unusual for properly aged A-286. It warrants a question to the supplier before the bolts enter production.

In this case, step 2 caught the 18°C offset. Our furnace control thermocouple had drifted, and because the customer asked to see the load thermocouple data rather than just the MTC summary, the problem was caught before the bolts entered their assembly line.

## What we fixed on our side

The incident triggered two changes in our heat-treat department:

1. **Load thermocouple data is now attached to every MTC as standard.** Not just for customers who ask. If you order aged superalloy bolts from us, your cert package includes the furnace chart with load TC data (setpoint, actual, and deviation) for the full aging cycle.

2. **Furnace uniformity surveys moved from quarterly to monthly for aging furnaces.** AMS 2750F requires quarterly surveys for Class 2 furnaces. We now survey aging furnaces monthly. The 18°C drift happened in the six weeks between quarterly surveys. Monthly surveys catch drift within two weeks.