Last updated 08/26/2004

Postulate

What is the actual temperature at which igniters burn? Do hand dipped igniters vary? How long do they burn?

Equipment

The data recorder is a DataPaq Tracker II with 10 channels. The data is accurate to +/- 1ºF and is (of course) NIST traceable. The thermocouples used were Type K, having chromel/alumel tips. They are asda" thick, having a asasd coating to protect against oxidation. This type of thermocouple has an operating range from 32ºF to 2,500ºF. The recorder is triggered by means of either timer, a programmed temperature or simply pushing the start button. (I chose the latter for this test.) The data was collected live onto an interfaced notebook computer.

Figure 1

Test set up

A reference was used at the beginning of the test and after the test. This ensures the proper correlation of each thermocouple and checks for any damage done during the test. The tips of all the thermocouples were bound together and heat was applied. This same method was used immediately following the test.

The assembled test specimens were placed in a fixture to secure them as well as organize them. The fixture was made by sandwiching two 1.5 ” x 7 ” 1/2' thick Kevlar panels. An ABC type fire extinguisher as well as a Halon fire extinguisher were ready in the wings. The assembly was ignited using a car battery and a switch.

Figure 2

Test specimens

The first test consisted of securing a thermocouple along side an igniter so that the thermocouple was touching the igniter halfway down the igniter. This round tested two each: Magnalite, Quick Burst and First Fire.You can see on the first graph that the Quick Bursts did not light right away. The switch was thrown again to light them. A problem popped up on the #4 thermocouple reading. A bad contact between the cold block and the DataPaq resulted in an invalid curve for #4. Two more Quick Bursts were tested to complete data points for this series.

I then wanted to know if there would be a difference in temperature from the surface of the igniter and the "core". I did a third test where I could test the heat with the thermocouple "in" the flame. Four specimens were made where thermocouple was secured to a nicrome igniter.(See figure 3) The assembly was then dipped in pyrogen, two with Magnalite and two with IgniterMan.

Figure 3

Data Collection

Figure 5 is the graph produced by the recorder during the first test. Two problems surfaced during this test. The first problem was that the Quick Bursts did not light the first time that the button was pushed. This can be seen by the time lag in the graph. When the other igniters went out the button was again pushed and the quick burst lit. The heat from the quick bursts caused a rise in the temp of the adjacent thermocouples. Looking at the graph, you'll see that the first fires had a secondary heating caused by this. The second problem was that a bad connection on thermocouple #4 caused an erroneous reading. The Quick Bursts were tested again and can be seen in figure 6.

Figure 4 Igniters lit	Figure 5 First Test	Figure 6 Retest of Quick bursts	Figure 7 Thermocouple dipped in Magnalite
Figure 8 Thermocouple dipped in Ignitorman	Figure 9 All Imbedded Thermocouples	Figure 10

Discussion

Some of the manufacturers report temperatures higher than than those that I recorded. It's not known from where these reported temperatures came. These may be theoretical temperatures at which the oxidation reaction takes place. In such a case, the temperature any distance away from this reaction would actually be less.

Another possibility is that the achieved temperature did not exist for long enough to transfer the heat from the reaction to the thermocouple. In this case, a longer burning igniter would impart more heat to the thermocouple.

A short experiment was conducted to attempt to determine which of these statements is true. Another type K thermocouple was used which has a faster reaction time. This one has the chromel/alumel wire significantly thinner, as seen in figure 10. This thermocouple was placed along side a Magnalite igniter.

In either case, the actual temperature touching the propellant grain is something lower than the reaction temperature. One statement that may be derived from the data is that the relative temperature of the igniters varies. This is true because whatever the "actual" temperature of the igniter is, the thermocouple recorded the temperature it "saw". A hotter thermocouple reading, be it due to a hotter burning igniter or a longer burning igniter, indicates that the given igniter is imparting more heat energy to the propellant grain. Thus, an igniter burning at 1,000ºF for 2 seconds will impart the same heat energy to the propellant grain as a 2,000ºF igniter burning for 1 second.

Theory Tested - New data

My theory that the heat did not exist long enough to transfer the actual heat to the thermocouple may have been correct. A different Type K thermocouple was used in another test. This type has leads and a tip that is the size of a human hair. These react very quickly. The new data warrants further testing. Please stand by....

New twist

Conclusion

Longer burning igniters impart more heat energy to the propellant grain.