I can see ideal diameter (DI) values being a useful predictor for the quench-and-temper processes. What effect does DI have on normalize (air quench) and temper of 4100- and 4300-series steels?

We have been trying to correlate DI with previous tests so that we can predict what temperature and speed to run new materials at. We would like to use DI to determine furnace conditions for normalizing and tempering rods that range in diameter from 0.75-1.5 inches (19-38 mm) and lengths from 2-25 feet (0.6-7.6 m).

The rods are currently run in a continuous furnace and go from normalizing at 1575°F to tempering between 1100-1250°F. The rods cool down to around 600°F on an exit table after normalizing.

My other question: Is DI the right constant to use to predict the outcome of mechanicals from new heat treating based on DI values and prior mechanicals of other tests?

Also, we currently compare DI values to average tensile and yield-strength results for a given hardening and tempering recipe. Currently, the mill does not provide mechanicals on the material, only DI and chemical analysis.

Finally, should we be using mechanicals instead to predict results?

You question is an interesting one. The data provided has allowed me to offer a few suggestions.

First, you need to do a reasonable amount of work to determine if your resulting metallurgical properties will correlate with DI. It will depend a great deal on what microstructure is developed in your process and the resultant mechanical properties. This will vary by steel type and DI value within that steel type.

The data gathered by you so far and sent to me is a good first step. You have provided a lot of useful information, and (if it hasn't already been done) someone now needs to run an in-depth analysis to determine if any (and what type of) statistical correlations exist. You also need to determine how each of the heat-treat sequences impact the part microstructure for each steel type you use so that you can construct your studies to compare apples to apples.

A cursory look at the data suggests that you might start by trying a study using one steel type with a given processing sequence (i.e. steel type per given heat-treat sequence). The expected metallurgical characteristics (e.g., hardness, microstructure, etc.) will be the same/similar and see what correlation you can get with DI. If it provides useful information, then continue to do that for each set of data for a given steel type/heat-treat-sequence combination. You should then have a correlation for each such combination and could predict for upcoming work using that combination.

Finally, the mill does not need to report properties since you are doing the heat treating as you will erase prior structure and properties. The mill can (typically at extra cost) report properties based on duplicating your heat-treat processes.

It is best to have a metallurgist do this work. Because it is not something that will be generated from tables, it will have to be empirically determined and correlated. It will depend on such issues as material specifications, processing parameters, equipment functionality and the resulting hardness and structure characteristics and the consistency thereof with respect to all the factors mentioned.