Process and Fixture Design | Highly customized
In flame brazing, measure the temperature of the workpiece, not the flame: infrared thermometers and thermal cameras cannot see a clean flame and are affected by the workpiece’s emissivity, so the error can be large; the industry practice is to fix thermocouples on the workpiece actually being heated and record the heating curve with a data logger.
The customer’s request was a single sentence: “Help us confirm the temperature of the flame.” But after Jiahon’s engineering team evaluated it, we found the question that really had to be answered was a different one. This article records how we turned a purchasing question of “which instrument to buy” back into a process question of “what to measure”.
Key points of this article
Flame temperature, workpiece temperature and heat input are three different quantities; what decides braze quality is the temperature history the workpiece actually goes through.
Whether long-wave infrared, ratio (two-color) or short-wave infrared, all have a common physical limit with clean, low-soot flames — this has nothing to do with the grade of the instrument.
No single instrument can meet all six conditions of a measurement system at once; the industry practice is to have the two functions of “continuous monitoring” and “periodic verification” side by side.
A request that looked simple
A customer manufacturing metal components came to Jiahon with a clear need: the production line has several flame brazing stations, and they wanted to confirm the flame temperature of each torch, considering fitting a non-contact temperature instrument at every station.
The need was reasonable and the starting point correct — what they wanted was joints that are consistent on every shift and at every station. But when our engineering team actually reviewed the site footage and the specifications of the instruments they planned to buy, we found that if they went ahead in that direction, the numbers they bought would very likely not answer the question they really wanted answered.
We did not rush to recommend another instrument; we took the problem apart first.
What the reflow oven teaches: a live display does not mean you know the workpiece temperature
The reflow oven in the electronics industry is a good reference. Every zone has a thermocouple, the temperature is displayed live on the panel, and it can be logged over a connection. Even so, customers still require suppliers to run actual products with attached thermocouple wires through the oven and measure the temperature curve the actual pads go through — known in the industry as profiling.
The reason is simple: the zone temperature reflects the state of the oven, while solder quality depends on the temperature the pads actually reach. Between the two lie the product’s heat capacity, board thickness, component shadowing and conveyor speed.
Displaying a temperature live and knowing the actual temperature of the workpiece are two different things.
Flame brazing is exactly the same; the variables in between just change to: distance, angle, heating time, heat drawn by the fixture, and the production cycle time.
The same principle has long been written into specifications at SMT and packaging plants. In our engineering data, an oven operating procedure from a packaging plant’s reliability lab states: the thermocouple must be fixed in the middle of the test sample with non-conductive adhesive, with as little adhesive around the junction as possible, to measure the true temperature; each oven’s temperature distribution is verified once a year with a 9-point recorder (1 point in the center of the chamber, 8 points in the corners, 5–15 cm from the walls); over-temperature protection is set at “operating temperature + 10℃” (example: 185℃ for 175℃). Measuring reflow oven temperature follows the same logic: fix the thermocouples on a profiling board, connect them to an oven temperature profiler and run it through once; that is how you know the curve the board actually went through.
First tell the three temperatures apart
Flame temperature: determined by the thermochemistry of the fuel type and the oxygen-fuel mixing ratio. With a fixed mixing ratio, it hardly changes with flow rate — in fact it is the stable one of the three.
Workpiece temperature: the temperature the joint actually reaches, the direct determining factor of braze quality.
Heat input: the heat transferred into the joint per unit time, affected by distance, angle, heating time, heat drawn by the fixture and so on — this is what really drifts day by day on the line.
So if the goal is joint consistency, the workpiece temperature is what really needs to be confirmed; the flame temperature is just an indirect indicator. This distinction decides every instrument choice that follows.
Why can’t instruments “see” a clean flame?
The principle of non-contact temperature measurement is to receive the radiation emitted by the target and convert it to a temperature. The three common types — long-wave infrared thermal cameras, ratio (two-color) pyrometers and short-wave infrared pyrometers — work on different principles, but the object measured must be something that glows by continuous thermal radiation. Metal surfaces are in this category; a bright yellow flame containing soot particles is too, because the soot glows incandescently.
But the blue flame common in industrial torches is not. It has very few soot particles, and its radiation is concentrated in the molecular emission bands of the combustion products (carbon dioxide at about 4.3 microns, water vapor at about 2.7 microns), which neither form continuous radiation nor mostly fall within the bands these instruments use. The result: the instrument “sees through” the flame and measures what is behind it — at most brazing sites, that is a whole hot hearth.
Figure 1 | The difference between two kinds of flame (schematic). The color of the flame is itself an indicator of soot content: yellow means it contains soot and can be measured by radiation instruments; blue means low soot, and what the instrument mainly receives is background radiation coming through the flame.
Even when aimed at the workpiece instead, each has conditions to confirm first: on a bright copper surface about 95% of the signal comes from reflection, and a hot background pours straight into the reading; during brazing, the surface also goes in turn through oxidation, flux coverage and molten filler flow, so the emissivity keeps changing. These limits are severe in long-wave infrared and milder in ratio and short-wave infrared, but none is immune.
One more thing is often overlooked: an instrument’s calibration certificate is issued under blackbody furnace conditions. A blackbody does not refer to color; it is an ideal radiator with an emissivity approaching 1. When the target changes to a semi-transparent gas or a highly reflective metal surface, the calibration conditions no longer hold — the accuracy of the instrument itself and the correctness of the measurement on site are two different things.
Six conditions for a measurement system
What Jiahon’s engineering team gave the customer was not “switch to another instrument”, but a set of criteria first. A temperature measurement system usable for process control should meet:
| Condition | Explanation |
|---|---|
| 1. Continuity | Data is obtained continuously during production, not just sampled at particular moments. |
| 2. Real time | Data is displayed in real time, so the effect of a parameter adjustment can be seen immediately. |
| 3. Repeatability and reproducibility | The variation of the measurement itself must be much smaller than the process tolerance for the data to be statistically meaningful. |
| 4. Truthfulness | What is measured must be the temperature the workpiece actually goes through, not the temperature of some other object. |
| 5. Recordability | Data can be logged and traced automatically, without relying on manual copying. |
| 6. Deployment and maintenance burden | The number of devices, setup conditions, and the resources for keeping calibration and traceability afterward. |
Checking against these six one by one shows: radiation instruments are generally strong on 1, 2 and 5 and weak on 3 and 4; contact measurement is just the opposite — it measures the workpiece directly with little variation, but is carried out periodically and is not continuous.
No single instrument can do it all — because these were always two functions
This is not a defect of any instrument. Look back at the reflow oven: the zone thermocouples handle continuity, real time and logging, and profiling handles truthfulness and reproducibility; the two exist side by side, and neither replaces the other. The right architecture for flame brazing is exactly the same:
Figure 2 | The division of work between the two functions. The suggested order of introduction is to establish the verification baseline first, then decide what the monitoring side needs — in the reverse order, the monitoring data has no basis for judging pass or fail.
Jiahon did not invent this; it is a long-established industry method
Flame brazing has its own specification, AWS C3.4M/C3.4 “Specification for Torch Brazing” (AWS product page), which covers equipment, materials, heating cycles and inspection requirements.
In heat treatment, AMS 2750 (aerospace) and CQI-9 (automotive) both list thermocouples fitted on the workpiece actually being heated as a standard sensor category, and use periodic verification (SAT, TUS) as the basis for ongoing process qualification.
Reflow profiling in electronics is the everyday application of the same logic.
On the electronics assembly side, IPC-7530B (guidelines for temperature profiling of mass soldering) likewise has a dedicated chapter on oven profiling with “thermocouples fixed on the board under test” (table of contents).
Even in the academic literature, non-contact temperature measurement studies of “flames impinging on metal surfaces” in the end still use thermocouple measurement as the verification baseline. In other words — whichever non-contact instrument you choose, it needs a contact baseline measurement once to confirm it reads correctly.
What happened next?
We compiled the analysis above into a technical report for the customer, including a frame-by-frame reading of the site footage, a comparison of the conditions under which each measurement method applies, and a list of technical questions suggested for clarification with the equipment suppliers. As a result, before buying any equipment, the customer first had a grasp of “what to measure, and how to know it is measured correctly” — the budget that might have gone into repeated investment in several instruments now had a clear basis for decisions.
This is how Jiahon works: do not rush to sell; define the problem correctly first.
Does your line also have temperature problems that cannot be measured accurately or explained clearly?
From choosing thermocouple wire, through multi-channel temperature logging, to building and interpreting process temperature curves, Jiahon’s engineering team can help you turn “we feel we measure it” into “we are sure we measure it right”.
Contact Jiahon’s engineering team
Further reading: TTK30 / GGK36 thermocouple wire — the specification differences explained →
Frequently asked questions
Q1 | Can an infrared thermometer or thermal camera measure flame temperature?
For a blue, low-soot flame, basically no. This kind of flame is nearly transparent in the bands the instruments use, and what the instrument mainly receives is background radiation coming through the flame. A bright yellow flame contains soot and can in principle be measured, but what is measured is an approximation of the soot temperature.
Q2 | What temperature should a brazing process actually monitor?
What has to be confirmed is the temperature history the workpiece (joint) actually goes through, not the flame temperature. The flame temperature is decided by the fuel and mixing ratio and is in fact a stable quantity; what really drifts is the heat transferred into the workpiece.
Q3 | How do I choose thermocouple wire?
It depends on the temperature range, the environment and the response speed needed; models and sheath materials differ greatly. See Jiahon’s thermocouple wire selection notes, or contact us directly, and our engineering team will make a suggestion based on your process conditions.