Updated: 11/7/18

Introduction

Having built my 1000lbf liquid-fueled rocket engine, I had always intended to ignite it with a solid cartridge. This has the simplicity of being very easy but makes multiple engine restarts slow and cumbersome, having to safe the test stand each time in order to reload another igniter cartridge.
However, since becoming a student, time nor funds are on my side to bring my engine to life anytime soon, so I needed an interim project, henceforth a spark torch igniter.

Aim

  1. Develop an igniter for the rocket engine.
  2. Gain more experience with coding and electronics design.
  3. Keep myself sane while studying a mechanical engineering degree.

Although this is not the first time I have made such an igniter, I have tried in the past with varying success rates, but it is only now that I have the time and patience to sit down and really nut it out. Plus it makes good study procrastination!.

Design

Having limited tooling and facilities, I decided to take the approach of a 3D printed igniter to make it as easy as possible. Shapeways do a bronze infused stainless steel which is heat resistant to 841°C (1114K) and well within the budget of a student such as myself. The choice of this material does require the igniter to be run fuel rich to keep the combustion temperature down as much as possible.

As this was going to be the first thing that I was going to have printed in metal I opted for an initial print in PLA. This allowed me to check threads, fits, and clearances etc. This proved very beneficial as some slight modifications of the original CAD design were needed. 1/8″ BSP threads were selected for the female fitting attachments and a 1/4-32 model airplane sparkplug for the spark.

The igniter is designed to run on a 75/25 Ethanol/water mix and gaseous oxygen and operate at a chamber pressure of 70psig. The choice of the fuel is the same as the engine.

Performance Specifications

Being the first iteration of 3D printed igniters, performance was low on the agenda, obtaining a reliable system and igniter were first and foremost.

  • Design total mass flow rate: 0.0099kg/s
  • Thrust (Calculated): 9.36N
  • Design chamber temperature: 992K
  • Design ISP: 96.38

Hardware

  • Arduino Uno
  • Arduino screw shield
  • 5V 4-channel relay module
  • x2 High-pressure solenoid valves
  • x3 1000psi pressure transducers
  • Rcexl CDI  electronic ignition box (1/4-32 plug)
  • Rcexl ignition test kit (to continually spark when power applied)
  • LM2596 Power Module
  • Converted fire extinguisher fuel tank.

A 12V battery provides power to the solenoids and Rcexl ignition box, although this is stepped down to 9V, as Rcexl recommend 12V max I decided to be on the safe side. The sensors and relay shield are powered from the 5V port on the arduino. This goes out to a 16 port plug and into the wiring loom to the respective parts. At the moment the Arduino requires the USB for power and communication, in the future I’ll convert this to RF to make it wireless.

Electronics setup

Control was initially through the Arduino serial monitor with command prompts, but my brother kindly spent some time in Visual Studio and made a GUI to enable control and have real time plotting. The GUI also produces a csv file with data which can be analysed at a later date.

Igniter GUI Screen Capture

Tips and Tricks

  • When choosing a solenoid make sure the solenoid orifice is larger than your injection orifice otherwise your solenoid will be the one setting the mass flow.
  • Depending on your setup and solenoids you may need a flyback diode in parallel with your solenoid to prevent voltage spikes, this I had to do with mine.
  • If you can afford it buy decent pressure sensors, although my $30 eBay ones worked fine, in my last few tests the oxidiser inlet sensor started to act erratically.
  • Printing injector orifices will take you a couple of goes and depends on multiple things, drill your own or buy a pre-made orifice like I did (unless you have cash to burn!)
  • Setup is everything, do whatever you can to quickly analyse and iterate tests. For instance, writing a script that would take all the raw data, make graphs, and show the results was extremely helpful. In the below-testing video, I was able to test, analyse data, make slight changes, and retest in less than 1 hour.
  • I used NGK ME-8 sparkplugs, these had a 1/4-32 thread and are more commonly found in 4-stroke model airplane engines. Make sure to buy a few to have spares on hand.

Testing

Hot Fire Testing – 4/7/2018 and 6/7/2018

I’ll post specific data and graphs here shortly.

Data Analysis – 6/6/2018

I spent some time developing a MATLAB script to post process all the data after a test. To verify this I threw in the data from my first hot fire to low and behold find I was way off the mark with the original performance specs calculated. I have since fixed this and found the igniter operated at lower performance than first thought, a few fixes are on the way before the next hot fire.

Water Flow testing – 12/4/2018

I managed to pick up some small orifice fittings that happened to thread perfectly into the bore of a 1/8 BSP nipple fitting, more here. They are within 4% of my calculated flow but are likely to be more accurate, as I only have a basic kitchen scale to measure the water mass. The fuel side was initially set up to swirl so some water retention could also be between the orifice and igniter body, I’ll drill a direct hole in line with the orifice and cover the swirl hole with metal putty.
Really happy with these orifices and setup, good to know all works after 4 months lying dormant!

Hot Fire Testing – ~End November 2017

After my first initial hot fire semi-success I was on a good high and happy everything worked. In order to show some friends how cool this was, I set up in the exact same manner as before and attempted another hot fire, this time I was unable to light the igniter.
The inlet pressures and valve timing were all exactly the same, the only difference I noted was the spark plug erosion, this was replaced and tried again. No ignition.

After talking to others I have a rough idea on how to proceed,

  • Switch from swirl fuel injection to 90deg with ox
  • 3D print (with SLS) smaller and more precise orifices that will press fit in my 1/8 nipple fittings
  • Future version, move inlets further away from spark plug to help prevent erosion or lengthen hole so does not protrude into chamber as much
  • I’ll start with the orifice fix and go from there.
  • Read more information here.

Hot Fire Testing – 14/11/2017

I achieved first hot fire today, I was happy how everything worked and was able to follow my setup and safety procedures without too much changes, so all in all a good day.

  • Chamber pressure (avg): 83.79 PSI
  • Fuel inlet/Feed (avg): 93.81/80 PSI
  • Oxygen inlet/Feed (avg): 102.05/140 PSI
  • Fuel mass flow rate: 0.00715 kg/s
  • Oxygen mass flow rate: 0.00926 kg/s
  • Mixture ratio: 1.29

The igniter ran too oxygen-rich in this test, in the next round of testing I will concentrate on getting the fuel inlet pressure right and then work on the oxygen feed until the desired ratio is correct.

Water Flow Testing – 1/11/2017

To validate the ethanol side of the igniter and to get the correct flow rate I carried out a series of water flow tests, capturing the water and equating this back to a mass flow rate. The tank pressure and igniter inlet pressure were also recorded to aid in flow validation.