Tempering of composites

Composite materials need to be heat treated to achieve best results. Both their final strength and heat-resistance increase. The reason is that epoxy resins do not polymerize completely at room temperature. Though a higher final strength is without question relevant, heat resitance is from a practical point of view much more important. Who wants to have a nicely shaped airplane deform in a hot summer day? Not me!

Heat resistance of a cold-hardening epoxy resin is usually about 30 °C over the curing temperature. This means that if the composite material is cured at 20 °C, the piece will not soften below 50 °C. The latter temperature is called the glass-transition temperature Tg, which depends on the degree of polymerization. Temperatures of over 50 °C are easily reached in the sun and some pieces can get very hot in motorized airplanes, such as the cowling. Even temperatures below but near to Tg should be avoided. Essentially, composites should be heat treated at the highest temperature they will be subjected to loads. The German Federal Aviation Authority (LBA) requires 54 °C for sailplanes and 72 °C for motorized airplanes, which means that pieces should be tempered at—or over—these temperatures.

Ok, so let’s heat  the piece up to 55 °C and everything is fine! Well, it is not that easy, as there are some pitfalls:

  1. During tempering the composite is heated over the original glass-transition temperature and deformation can take place.
  2. Composites can deform heterogeneously when heated.
  3. Tighly regulated heating of cubic meters of space is challenging.

The solution to the first problem is to heat up slowly and if possible, to temper inside the mold. Consequently, this means that the mold itself has to resist the temperature. Slow heat production provides the remaining resin components enough time to react resulting in a continuously rising glass-transition temperature. In other words: if you do it right, you never reach Tg. Consensus is that the heating rate should stay below 10 °C per hour. The second point is solved by distributing heat as good as possible and by employing quasi-isotropic laminates in the mold. Quasi-isotropic laminates tend to deform homogeneously (at least in the plane of the laminate). And the solution to the third point is at hand: power. However, you need to account for points one and two! Power, thus, has to be regulated tighly, or else, the composite might be spoiled.


Today, I’d like to present my solution to these problems. There are industrial solutions, which are too expensive to be used by private persons. RC-Model builders use boxes of isolating material. They heat by turning light bulbs on, and regulate the heating rate by the amount of lighted bulbs. Other people heat using old electric ovens or with gas heaters. They usually regulate the rate and final temperature by allowing heat to escape up to a certain degree. An approach that is energy-intensive and needs a careful hand to setup correctly.

I decided to use a microcontroller which drives an electric heating source through a solid state relay :


The board is an Adafruit Adalogger based on a Cortex M0+ microcontroller. It has a OLED display and a SD card reader/writer. The temperature profile is written on the SD card and the current state is easily read from the display.

As a heating source I decided to use heating foil, which is usually installed in room ceilings:


This foil provides about 200 W per square meter. I started by building a small prototype to desgin the software:



Basically, the heating foil is driven by a pulse-width modulated signal. The microcontroller decides with the current temperature the duty cycle and, thus, the effective heating power.

Regulating a constant temperature and a heating rate is more challenging than I thought at the beginning. It is not difficult when variation of several degrees is allowed, such as in a fridge. Here, however, I wanted to have tighter control. The temperature sensor I use (DS18B20) provides a discrete temperature, something that controllers do not like at all. So I had to come up with a solution to that…

A test run of the prototype showed that my ideas work quite well. Depending on the heating rate, the control error is small below 0.3 °C. This is certainly better than regulating heat escape by hand!


I built a larger box to test if everything works well on a larger scale. It is large enough to fit the center section:


The test is running since a couple of hours and I am eagerly waiting for the results. The room is cold (about 6 °C) and even so it achieved yesterday night a temperature of over 40 °C. The power should be enough to reach 55 °C when the room has 20°C or more!



Kommentar verfassen

Trage deine Daten unten ein oder klicke ein Icon um dich einzuloggen:


Du kommentierst mit Deinem WordPress.com-Konto. Abmelden / Ändern )


Du kommentierst mit Deinem Twitter-Konto. Abmelden / Ändern )


Du kommentierst mit Deinem Facebook-Konto. Abmelden / Ändern )

Google+ Foto

Du kommentierst mit Deinem Google+-Konto. Abmelden / Ändern )

Verbinde mit %s