In the two major global markets (aerospace and automotive) served by the heat-treatment industry, quality has always been a key driving force. To many heat treaters, the concept of quality is a double-edged sword. On the one hand, it’s a cost to be borne to compete in these markets and, on the other, a competitive advantage to those companies able to achieve higher levels of accreditation and product than their peers.

Regulatory standards are driving process control and quality higher up the agenda of the users of heat-treated products. Advances in furnace control are enabling heat treaters to meet these demands in a cost-efficient manner. As Henry Ford, the grandfather of today’s automotive industry, once observed: “Quality means doing it right when no one is looking.”

Fig. 1. Schematic of the 3G Plus system

Contemporary Furnace Control

The ability to measure and control all parameters in a process enables a company to assure the quality of its products. The gas-carburizing process is a cornerstone in the modern production of automobiles and aircraft, as well as many other industries. However, the critical process of dissolving carbon atoms into the surface of the steel component takes place in a furnace when no one is looking.

Contemporary control systems allow the user to control many of the key process variables, but they work on the principle that the levels of certain unmeasured furnace gases, such as carbon monoxide and methane, remain constant throughout the process. The nature of the carburizing process means that the furnace atmosphere will vary during the process, and the level of uncracked methane in the furnace does have an effect on the end result. This can be overcome by the use of a “correction” factor, a constant calculated when the furnace is commissioned and applied to furnace recipes, which enables the user to produce the required case depth.

Most furnaces run many different loads over different cycles, however, and a single constant is unable to deal with the variations of multiple variables in different circumstances. This makes the rate of carburizing and the resulting case depth more difficult to predict, and it can lead to poor quality and scrap where case depths do not meet customer requirements.

Carbon-Probe Limitations
A carbon (oxygen) probe has high repeatability and is a good comparator with rapid response, but the accuracy is overstated for a variety of reasons:

1. High millivolt outputs
  • The probe sheath-material construction (nickel alloy/platinum) can result in the catalytic breakdown of free methane into CO and H2, resulting in high millivolt outputs in high free-methane atmospheres.
  • Sooting – ineffective carbon-probe burnoffs (probe cleaning) will result in high local levels of carbon potential (Cp).
2. Low millivolt outputs
  • Contamination of reference air
  • Sensor leakage

Fig. 2. Effect of free methane on carbon potential

How Errors Can Creep into a Process

Non-Equilibrium Atmospheres
All manufacturers’ carbon-potential calculations assume equilibrium gas conditions; that is, the main furnace gas reactions are in balance. These conditions rarely exist in a furnace and will take many hours to come into equilibrium.

For example, it is assumed that the CO content in endothermic carrier gas, produced from methane, is 20%. At the start of a heat-treatment cycle, the CO content may fall to 16%, recovering to 20% over an hour or so. The calculated Cpassumes a constant value for CO content in the furnace atmosphere. Because it varies, there is a need to use a correction factor in the Cpcalculation to compensate, and this will be an average value over the total cycle.

Carrier-Gas Variations
The CO content in the furnace carrier gas may vary depending on the source. For endothermic gas generators (typically 20%CO), the condition of the generator catalyst and the air:gas ratios are factors.

For nitrogen/methanol systems (typically 16-20%CO), ratio deviations, methanol impurities, nitrogen bubbles in methanol, condition of vaporizer and low furnace temperature (poor cracking) all affect the CO content.

The Effect of Free Methane
In an atmosphere with significant free methane, carburizing takes place, but as the level of free methane increases, it has a dilution effect on the atmosphere, thereby reducing the overall %CO. Without measuring and taking into account the effects of free methane in the carbon-potential calculation, the real carbon potential is not known.

A New Vision of Your Process

Newer carburizing technology takes the quality assurance of gas-carburizing furnaces to a new level. Designed to eliminate rework and scrap by giving the furnace-control system the most comprehensive vision of the furnace atmosphere, such systems measure the levels of CO, CO2and CH4in the furnace atmosphere. It uses these to apply a compensation to the oxygen-probe reading in order to calculate the true instantaneous carbon potential in the furnace.

These analyzers are often sold as an enhancement to an atmosphere-control-solution software, serving as a single-point system dedicated to one furnace. In such cases, the analyzer extracts the furnace atmosphere, using its internal sample pump, via the sample line, filters and flowmeter. The furnace atmosphere content of CO, CO2and CH4is analyzed by three individual NDIR analyzers (one for each gas). The system then calculates the “IR Cp,” based on the values of the gases CO, CO2, CH4, and temperature to accurately determine the furnace atmosphere carbon potential. The analyzer then calculates a correction factor, based on the IR Cp, that is required to make the carbon-probe Cp equal IR Cp. When the system is in active mode, the calculated correction factor is applied to the carbon-probe Cpcalculation.

The result: The carbon potential now reads the same as the calculated IR carbon potential. The carbon probe may be sooted, or even failing, but the system will compensate. The absolute accuracy of the carbon probe becomes unimportant. The carbon probe handles the second-to-second Cpcontrol loop, its calculated Cpbeing regularly updated by the system.

The type of correction factor employed is user-selectable based on one of the following:
  • Process factor
  • CO factor
  • Probe mV offset
One example of this technology is the Eurotherm 3G Plus system, which is intended as an enhancement to the company’s LIN-based atmosphere-control solution based on Eycon and T2550 controllers. Trials have shown that a furnace controlled using this system is three times more accurate than one regulated with an oxygen probe alone, with a tolerance of ±0.06% of Cp. Integrating the three-gas analyzer within the furnace control system enables highly repeatable treatments with quality built into the process. In an industry where quality is paramount, the 3G Plus adds a new dimension of accuracy and repeatability to process control.IH

For more information:Peter Sherwin is the business development manager, heat treatment for Eurotherm, 741-f Miller Drive SE, Leesburg, VA 20175; tel: 571-246-3809; e-mail: Peter is interviewed by IH publisher Doug Glenn in the Eurotherm video on IHTV at

Additional related information may be found by searching for these (and other) key words/terms via BNP Media SEARCH at furnace control, gas carburizing, oxygen probe, carbon-probe burnoff, carbon potential, endothermic gas, nitrogen/methanol, NDIR