Achieving the elimination of oxidation during thermal treatment has driven heat treaters for decades and resulted in a wide variety of approaches. The obvious method is to flow an inert gas such as nitrogen into the furnace in order to drive out both air and moisture. By itself, however, this technique is inadequate.
The zirconia carbon sensor has been used for nearly three decades to control the carbon potential in many carburizing applications. Today’s best of the web article examines the use of the zirconia carbon sensor in a variety of annealing and special treatment applications and considers how the sensor millivolt output is preferred because it relates directly (not empirically) to the free oxygen concentration in the surrounding environment.
An Excerpt:
“While it is desirable to avoid oxidation during thermal treatment, the achievement of adequate control using one of the ‘getter’ gases requires that the sensor millivolts achieved be established at some value higher than the vee formed by the iron reaction at temperatures below 1375ºF and the carbon reaction above that temperature. The vee will demonstrate the lower limit, but the practical level should be established by evaluation of product quality, getter cost and possible sooting. The appropriate level will be limited by such things as furnace leaks, atmosphere agitation, work porosity, time of treatment, etc.”
What are advanced management systems and how does deep integrative system management software help automotive heat treaters improve processes while saving on time and unnecessary expenses? Explore the future of software technology for the management of heat treating operations in this Technical Tuesday by Sefi Grossman, founder and CEO of CombustionOS.
The heat treating industry is on the brink of a technological transformation. Just as the momentous adoption of websites and emails transformed the nature of work for manufacturers, the advanced software systems are thrusting us into a new era of simplicity, automation, and deep integrations.
This article explores how advanced systems — an application of ERP (enterprise resource planning) and MES (manufacturing execution systems) combined with the power of AI — is revolutionizing facility operations, enhancing quality, efficiency, and profitability.
What Are Advanced Systems?
Advanced systems simplify, streamline, and automate operations by lifting the data burden off of plant personnel. While most existing systems focus on the part inventory workflow, more advanced systems go beyond by directly integrating into the heat treat process to track at bin/tray/tree level.
This requires real-time scheduling control, barcode scanning, digitizing recipe and process (no more paper), and direct sensor/PLC integration. Because of its critical nature, an advanced system is most likely an on-premise and cloud “hybrid solution” that is not crippled by internet connectivity issues. This allows it to still utilize rapidly evolving cloud systems that provide external services like messaging, big data storage, and AI to name a few.
Precise Processing
Figure 1. CombustionOS developers spend extensive time with operators and plant managers to create interfaces that are intuitive and easy to use. Pictured is access to job data stats from a mobile device being used outside of the manufacturing plant.
Repeatable, accurate methods to ensure optimal time, temperature, and atmosphere of the decided heat treatment processes are possible with advanced systems.
Utilizing existing sensors and hardware interfaces, data is collected in short intervals, transformed into meaningful data formats, and stored in a database. Network technologies such as HTTP, Modbus, and other analog to AI technologies make this possible with minimum additional hardware. The data is managed locally on the facility network, and synchronized with cloud services for further processing, analysis, and long-term history storage.
With a close monitoring of all these variables, facilities can tighten acceptable specification ranges. Deep integration with equipment ensures that data flows seamlessly from sensors and devices to the central system.
This real-time data collection and processing enables facilities to monitor operations continuously and make informed decisions quickly. For example, integrating data from temperature sensors, pressure gauges, and other monitoring devices ensures that all critical parameters are tracked and managed effectively. Additionally, if a temperature reading deviates from the acceptable range, the system can immediately alert the relevant personnel, allowing them to take corrective action before it becomes a critical issue.
In addition to quality assurance, integrated artificial intelligence tools optimize job scheduling. Unlike traditional date/time calendar methods, AI systems predict job completion times based on real-time process data. This is particularly useful for roller furnace setups, where continuous processing occurs, but it is also beneficial for batch furnaces. Optimized scheduling improves resource allocation and operational efficiency, ensuring that jobs are completed on time and to the required specifications. The difference between a “calculation algorithm” and AI is that, with AI, you do not have to pre-program it. It automatically learns and adjusts for known variability in your hardware and even the personnel that are operating the equipment.
Finally, the automation of these systems captures and records all necessary information accurately. This reduces the risk of non-compliance, improving the overall quality of the final product. For example, a Detroit-based heat treating facility reported that accessing real time data to ensure compliance with industry standards has allowed them to spend 40% less time on documentation tasks.
Figure 2. Having increased control over the process gives more peace of mind to operators that components perform as needed.
Alleviating Burden on Maintenance and Inventory
Predictive maintenance is one of the most significant applications of AI in the heat treating industry. Traditional maintenance schedules are often based on fixed intervals, which can lead to unnecessary downtime or unexpected failures. AI driven predictive maintenance, on the other hand, uses real-time data to determine the optimal times for maintenance activities. This approach not only reduces downtime but also extends the lifespan of equipment.
A Detroit-based heat treating facility implemented an AI-driven predictive maintenance system (PMs) and saw a 25% reduction in equipment downtime. By analyzing data from critical parts, inventory, process tracking history, and various sensors, the AI system could predict when components were likely to fail, allowing the maintenance team to inspect and address issues proactively beyond their standard PMs. This not only improved operational efficiency, but also saved significant costs associated with emergency repairs and unplanned downtime.
Additionally, the integration of QR codes for inventory and process tracking enables quick and accurate data entry compared to manual logging. For instance, when racking parts out of bins, operators can simply scan QR codes, which automatically update the system with the relevant information. This not only speeds up the process but also minimizes the chances of human error.
Reducing Operational Costs
The adoption of advanced ERP and MES systems has led to substantial cost savings for many facilities. These systems reduce operational costs through the implicit automated integrations that technologies like CombustionOS bring. Here are just a few ways that operational costs have been cut:
Decreasing shipping and receiving management from three to just one employee
Minimizing rework costs by timely process alerts
Reducing personnel by replacing constant manual oversight with accurate, digital tracking systems
Lowering administrative costs by utilizing a more efficient and accurate invoice automation platform
Case Study: A client reported comprehensive cost savings, including a 20% reduction in shipping and receiving time, fewer logistics and furnace operators needed, a 33% decrease in rework costs, a 15% savings in maintenance costs, and a 25% reduction in accounting overhead. These efficiencies translate into substantial payroll savings and improved profitability.
How To Implement
Figure 3. When racking parts out of bins, operators can simply scan QR codes, which automatically update the system with the relevant information.
One of the most significant advancements in heat treating technology is the deep integration with various equipment types. Unlike traditional ERP systems, which often lack true integration, advanced systems work backwards from equipment data, building ERP functionalities around this integration to ensure seamless and accurate data flow.
First, there are advanced systems that can handle data from both digital and analog sensors. So, for heat treaters who are juggling a variety of sensors and systems, looking for an integrative advanced system that has adaptability will ensure compatibility with existing equipment while keeping an eye on cost. Facilities can continue using their current equipment while benefiting from advanced monitoring and control capabilities.
Second, advanced ERP/MES systems can take collaboration with multiple vendors. Rather than uproot current systems and relationships, work with an advanced systems provider who is able to collaborate with other software and systems. Advanced ERP/MES systems provide comprehensive solutions that include deep equipment integration and full ERP functionalities. This approach reduces the complexity and cost of integration, ensuring that all components work together seamlessly.
Key Applications
Most operations in a heat treat department will benefit from advanced systems due to the time-saving automations that the system integrates. But many heat treaters are looking to adapt and integrate older systems and often more complex designs, like roller hearth furnaces. Here are some steps that experts will take to guide you through to make the digital integration smooth and effective:
First, it is important to understand you don’t need to boil the ocean. Starting with a more advanced inventory tracking system that employs barcodes can set the underpinnings for a more integrated system while providing immediate benefits to your logistics.
Then, it is also key to get a deep understanding of your current process and map out your operational workflow. Using a flowchart program helps visualize the process to make sure all stakeholders are on the same page.
Some aspects of your current process are probably outdated (perhaps created by someone who is no longer at the company), while others are key to the core of how you operate. Understanding the difference is crucial to make sure you unlock potential automation without disturbing your core process and flow.
You’ll then need to prepare every required form, document, chart etc. that you use in the operation. For process control, recipes, and lab testing, provide many parts/iterations to capture the complexity.
Finally, take inventory of any existing digital systems you have adopted, like inventory tracking, spreadsheets, or custom software. The existing system network, including servers, Wi-Fi setup, and hardware (PCs, printers, scanners, etc.) will be utilized as much as possible in the transition to reduce the need to purchase and set up different equipment.
The future will require constant innovations and thoughtful leveraging of increasingly advanced systems. Unlike static, homegrown, or “pieced together” solutions, the most advanced systems are constantly updated with new features, ensuring they remain at the cutting edge of technology. Engaging directly with plant personnel to understand their needs and challenges allows systems like CombustionOS to evolve and improve continuously.
The heat treating industry is on the cusp of a technological transformation, driven by advancements in ERP, MES, and AI. These technologies offer the potential to enhance quality, efficiency, and profitability, making them essential for the future of manufacturing. By embracing automation, integrating advanced AI capabilities, and committing to continuous innovation, the industry can achieve new levels of operational excellence.
About the Author:
Sefi Grossman Founder & CEO CombustionOS Source: Author
Sefi Grossman has been at the forefront of technology revolutions for the past two decades and has been leading the technology company CombustionOS for nearly seven years.
Carburizing. It must happen sometimes, and if your heat treat division truly understands the impact of the atmosphere, more power to them. In this article by Jim Oakes of Super Systems, you will learn how seeing simulated data with real-time data can help you predict the amount of carbon available to the steel surface.
An excerpt:
“It is important to understand the model and specific variations caused by temperature, furnaces, agitation, fixturing, and part composition. Variations include alloying effects on the diffusion modeling based on certain alloy components, such as chromium and nickel.”
When processing critical components, heat treaters value and demand precision in every step of the process — from the recipe to data collection — for the sake of accurate performance of the furnace, life expectancy of all equipment, as well as satisfactory delivery of a reliable part for the customer.
So what’s the obstacle to achieving those goals? Gunther Braus of dibalog GmbH/dibalog USA Inc. says, “The general problem is the human.” Indeed, the need to remove the variable of human fallibility plays a significant role in the search and development of equipment that could sense, read, and record data separate from any input from the operator. “As long there is a manual record of values there is the potential failure,” adds Braus.
Now, as part of the quest for precision, particularly in the automotive and aerospace industries, many control system requirements are driven by the need to prove process compliance to specified industry standards like CQI-9 and AMS 2750. These standards allow for and frequently require digital data records and digital proof of instrumentation precision.
With this in mind, Heat Treat Today asked six heat treat industry experts a controls-related question. Heat TreatControl Panel will be a periodic feature so if you have a control-related question you’d like addressed, please email it to Editor@HeatTreatToday.com and we’ll put your question to our control panel.
Q: As a heat treat industry control expert, what do you see as some of the best practices when it comes to digital data collection and storage and/or validation of instrumentation precision?
We thank those who responded: Andrew Bassett of Aerospace Testing & Pyrometry, Inc.; Gunther Braus, dibalog GmbH/dibalog USA Inc; Jim Oakes of Super Systems, Inc; Jason Schulze, Conrad Kascik Instrument Systems, Inc.; Peter Sherwin, Eurotherm by Schneider Electric; and Nathan Wright of C3Data.
Calibration and Collection
Jim Oakes (Super Systems Inc.) starts us off with an overview of the equipment review process, the crucial component of instrument calibration, and digital data collection:
“Industry best practices are driven by standards defined by the company and customers they serve. Both the automotive and aerospace industries have a set of standards which are driven through self-assessments and periodic audits. Instrument precision is defined by the equipment’s use and is required to be checked during calibrations. The frequency of these calibration depends on the instrument and what kind of parts and processes it is responsible for.
The equipment used for these processes can be defined as field test instrumentation, controllers, and recording equipment. Calibration is required with a NIST-traceable instrument that has specific accuracy and error requirements. Before- and post-calibration readings are required (commonly identified as “as found” and “as left” recordings). During calibration, a sensitivity check is required on equipment and is recorded as pass/fail. The periodic calibration procedure is carried out not only on test equipment but also on control and recording equipment, to ensure instrument precision.
Digital data collection is a broad term with many approaches in heat treatment. As mentioned, requirements are driven by industry standards such as CQI-9 and AMS 2750. Specifically when it comes to digital data collection, electronic data must be validated for precision; checked; and calibrated periodically as defined by internal procedures or customer standards. Data must be protected from alteration, and have specific accuracy and precision. Best practice tends to be plant wide systems that cover the electronic datalogging that promotes ease of access to current and historical data allowing use for quality, operational, and maintenance personnel. Best practices in many cases are defined by the standards within each company, but the hard requirements are often the AMS 2750 and CQI-9 requirements for digital data storage.”
Industry Guidelines and Requirements
Andrew Bassett (Aerospace Testing & Pyrometry) has provided us with a reminder of the industry guidelines for aerospace manufacturing (via AMS-2750E, paragraph 3.2.7.1 – 3.2.7.1.5)
The system must create electronic records that cannot be altered without detection.
The system software and playback utilities shall provide a means of examining and/or compiling the record data, but shall not provide any means for altering the source data.
The system shall provide the ability to generate accurate and complete copies of records in both human readable and electronic form suitable for inspection, review, and copying.
The system shall be capable of providing evidence the record was reviewed – such as by recording an electronic review, or a method of printing the record for a physical marking indicating review.
The system shall support protection, retention, and retrieval of accurate records throughout the record retention period. Ensure that the hardware and or software shall operate throughout the retention period as specified in paragraph 3.7.
The system shall provide methods (e.g., passwords) to limit system access to only individuals whose authorization is documented.
“One of the biggest issues I see with one of these requirements will be point 5,” says Bassett. “The requirement is to be able to review these records throughout the retention period, which in some instances is indefinite. I always recommend to clients who may be upgrading or purchasing new digital systems that they should consider keeping a spare system in place to be able to satisfy this requirement. Who knows — today we are working on Windows 10, but in 50 years, will our successor be able to go back and review heat treat data when everything is run on Windows 28?”
“This is a topic that yields great discussions,” adds Jason Schulze (Conrad Kascik). He directs us to a challenge he sees from time to time.
Within the Nadcap AC7102/8 checklist, there is this question: “Do recorder printing and chart speeds meet the requirements of AMS 2750E Table 5 or more stringent customer requirements?” This correlates with AMS2750E, page 12, paragraph 3.2.1.1.2 “Process Recorder Print and Chart Speeds shall be in accordance with Table 5”.
“To ensure the proper use of an electronic data acquisition unit used on furnaces and ovens, these requirements must be understood,” continues Schulze. “Because this system is electronic, it should be designated a digital instrument and not an analog instrument. In doing so, this helps determine what requirements apply in Table 5. The only remaining requirement in Table 5 for digital instruments is ‘Print intervals shall be a minimum of 6 times during each time at temperature cycle. Print intervals shall not exceed 15 minutes.’
With this in mind, it is important to realize that, if your time at temperature cycles are short cycles (such as vacuum braze cycles), the sample rate of data collection may need to be adjusted to ensure it is recorded 6 times during the cycle.
As an example, if the shortest cycle processed is 4 minutes at temperature, a sample rate of every 60 seconds would not conform to AMS2750E because, in theory, the maximum amount of recordings would be 4 times during the time at soak. Now, if the sample rate was modified to every 30 seconds, this would allow ~8 recordings during the time at soak, which then would be conforming to AMS2750E.
Within the realm of electronic data acquisition on furnaces/ovens, this seems to be a frequent challenge for suppliers.”
A Critical Variable: Process Temperature
Nathan Wright (C3Data) agrees and zeroes in on process temperature as a critical variable to be measured:
“No matter the heat-treating process being carried out, complying with AMS-2750 and/or CQI-9 requires that the heat treater measure, record, and control several different variables. One of the more common variables that must be measured, recorded, and controlled is process temperature.
Measuring process temperatures requires the use of a precise measurement system (Figure-1 below), and the accuracy of said measurement system must be periodically validated to ensure its ongoing compliance.”
“The validation process is carried out through a series of pyrometric tests (Instrument Calibration and SAT), and historically these validation processes are highly error-prone.
In order to help ensure process instrumentation, process temperatures, and any other variable that impacts quality is properly validated it is good practice to begin automating compliance processes whenever and wherever possible. C3 Data helps automate all furnace compliance processes using software.”
A “Standard” Mindset
Gunther Braus (dibalog) chimes back in with some pertinent wisdom: “It is not sufficient only to record, you must live the standards like CQI-9, AMS, Nadcap or even your own standard you have set up, so you must survey the data. However, in the old times, there was a phrase: the one who measures, measures crap. In the end, it is all about surveillance of the captured data.
Where you store the data is a question of philosophy: personally, I prefer local storage in-house. Yes, we all talk about IOT, etc., and I do not want to start a discussion about security; it is more about accessing the data. No internet, no data. So simple. We are overly dependent upon cloud usage on the internet.
The automation of the instrumentation precision is so much effort in terms of automated communication between testing device and controller, from my point of view we are not there yet.”
A Look at the Standards In and Outside the Industry
The aim is to record the true process temperature seen by the components being treated. However, there are many practical factors that can alter the accuracy of the reading. From the position of the thermocouple (TC), the TC accuracy (over time), suitability of the lead or extension wire, issues with CJC errors and instrument accuracy as well as electrical noise impacting the stability of the reading.
The standards do a good job to help by prescribing the location of TC, accuracies required for both TC and instrument, and frequent checks over time through TUS and SAT checks but note the specification requirements are maximum “errors”. And if you truly want to reach world-class levels of process control and reap the inherent benefits of better productivity and quality, you should aim to be well inside those tolerances allowed.
With 30yrs+ of data required to be stored (in certain cases, particularly aerospace), there should be some thought as to how and what form this should be stored in. There are many more options of storage when the data is in digital format.
Paper is very costly to store and protect.
The virgin data file should be secure and tamper-resistant and identical copies made for backup purposes held offsite.
The use of FTP is becoming more common to move files automatically from the instrument to a local server (with its own backup procedures to ensure redundant records in case of disaster).
Regular checks should be made to examine the availability and integrity of these electronic records.
Control and Data Instrument suppliers should ideally have many years of supplying instrument digital records with systems that can access even the earliest of data record formats.
We also look outside of the heat treat standards for truly best practices. The FDA regulation 21CFRPart11 and associated GAMP Good Automated Manufacturing Practice have been extended with the new document “Data Integrity and Compliance with Drug cGMP, Questions and Answers, Guidance for Industry”. These updates leverage A.L.C.O.A to describe the key principles around electronic records (see below). This industry is also leading the requirement for sFTP a more secure format of the FTP protocol.
Heat TreatToday will run this column regularly featuring questions posed to and answered by industry experts about controls. If you have a question about controls and/or data as it pertains to heat treating, please submit it to doug@heattreattoday.com or editor@heattreattoday.com.
Controlling process temperature with accuracy and without extensive operator involvement is a crucial task in the heat treat shop and calls for the use of a temperature controller, which compares the actual temperature to the desired control temperature, also known as the setpoint, and provides an output to a control element. This comparative process relies upon an algorithm, the most commonly used and accepted in the furnace industry being the PID, or Proportional-Integral-Derivative, control.
“This popular controller is used because of its robust performance in a wide range of operating conditions and simplicity of function once understood by the processing operator,” writes Real J. Fradette, a Senior Technical Consultant with Solar Atmospheres, Inc, and the author of “Understanding PID Temperature Control as Applied to Vacuum Furnace Performance” (with William R. Jones, CEO, Solar Atmospheres, Inc, contributing).
The PID algorithm consists of three basic components, Proportional, Integral, and Derivative which are varied to get optimal response. If we were to observe the temperature of the furnace during a heating cycle it would be rare to find the temperature reading to be exactly at set point temperature. The temperature would vary above and below the set point most of the time. What we are concerned about is the rate and amount of variation. This is where PID is applied. ~ Fradette
In this week’s Technical Tuesday, we direct our readers to Fradette’s article at Solar Manufacturing’s website where he and Jones cover the following on PID temperature controllers:
Definitions, e.g., Closed Loop System; Proportional (GAIN); Integral (RESET); and Derivative (RATE)
Actual operation of a PID temperature controller, including understanding PID dimensions and values; and general rules for manually adjusting PID
and digital data collection:
Andrew Bassett (
“This is a topic that yields great discussions,” adds Jason Schulze (
Gunther Braus (
