De KNSB-2 is the second test with this small rocket motor. Although the use of aluminum tubes generated problems in the first test (see report static test KNSB58-1) it was decided to use the same type of tubes again. Bert Kimpe suggested replacing the silicon liner used in the first test by parabond. A somewhat thicker liner was used (about 1 mm).

The test was conducted on November 18, 2006.

The motor has the following characteristics:

For the preparation of the 3 propellant segments, 2 types of KNO₃ were used: 2 modules used KNO₃ Greenhouse Grade (of Haifa) and 1 module used KNO₃ from MERCK. The last type was mixed for 5s in a coffee mill. The KNO₃ from Haifa was not milled. This type of KNO₃ has the drawback that sorbitol decomposes during melting. To overcome this problem 5% of sorbitol was replaced by ascorbineacid (as recommended by Richard Nakka).

For the melting of the propellant use was made of an adapted melting system, constructed by Leopold Clauwaert. This system can be closed completely to allow vacuum creation. It can be completely submerged in the oil bath and as such assures a quasi homogeneous temperature. The powder constituents of the propellant were mixed before being added to the melting chamber, after which the melting system was submerged in the preheated oil.

Since the melting system was not provided with enough space to contain all propellant as powder, it was necessary to open the system after a first melting period of about 45 minutes and to add the remaining propellant powder to the system. The total melting time was between 90 and 110 minutes. Only limited discoloration of the propellant was observed. .

The propellant based on GG KNO₃ needed about 24 hours before being sufficiently strong, while it took only about 10 hours for the propellant with KNO3 from MERCK to reach more or less the same value.

The 2 first batches where processed between 135 en 143°C. The last batch (with KNO3 from Merck) was treated at an oil temperature of 121°C .

At the end of the melting process the propellant was subjected to a reduced air pressure (it was not possible to produce high vacuum as was intended).

The different segments were placed in the motor in the following way:

- All segments with the smallest inside diameter facing the nozzle

- The segment with the Merck KNO3 closest to the nozzle.

To assure ignition the same igniter as for the KNSB58-1 was used. This is a steel tube of 10mm in diameter provided with 36 holes of 1 mm in diameter spiralyzed over the total length. The tube was filled with 60 g of zinc-sulfur powder (3/1). The igniter was attached to the diaphragm such that it would be ejected at diaphragm burst.

Discussion :

As can be seen from the thrust curve, the motor was very quickly ignited. The simulation indicates that between 70% to 80% of the total propellant surface is ignited at time zero.

The total duration is about 3.5s which is significant longer than KNSB58-1 (about 2.5s). The difference may be due to the usage of the GG KNO₃ in two segments.

This time the aluminum tubes did very well. This is without any doubt thanks to the use of parabond. As already explained also the thickness of the polymer was higher than in the previous test. Also the outside of the aluminum tubes are still intact (this time no grease was applied on the outer surface of the aluminum tubes).

Figure 2: the complete motor

Figure 3: ignition system

Figure 4: mounting igniter and diaphragm

As already mentioned a 70 to 80% ignition of the propellant surface gave the best fit in de computer simulation. To cover the period from the start to the top of the thrust curve the ignition of the remaining propellant surface was taken as a linear relation with time ( a regressive relation would have been more realistic). The simulation program obliges to reduce the burning rate of the propellant (at zero erosion) at the actual pressures, with about 20%-30%, depending on the used erosion parameters in order to cope with the measured thrust. This is not new and was already found in the BEM motors of the SS2S project.

The tail of the thrust curve does not correspond well with the simulation. This is probably do to the fact that in the computer simulation a threshold value for the mass rate is used before erosion can occur. All propellant before this point is supposed to burn without erosion and creates a plateau in the thrust curve. This may not be the reality.

Fig. 5: thrust curve of KNBS58-2

Fig. 6& 7: inside of the aluminum tubes after test

The measured Isp is 89s. This is in agreement with the computersimulation.

A total of 50 g of remains were measurted after the test.

Figure 8: simulation of the thrust curve

Figure 9: photo of the test

The project is a collaboration between Tony Vyverman and Leopold Clauwaert. Also Bert Kimpe made important contributions to the project.