DukeAERO — Liquid Engine Ignition Hardware

I’m developing two ignition assemblies to support initial testing of DukeAERO’s liquid engine: a nozzle-mounted pyrotechnic igniter and a propane-torch assembly for future upright tests. I’m responsible for the igniter designs and work with teammates to integrate them with the engine and test setup.

The nozzle-mounted assembly has been built and checked for fit and mechanical retention through bench loading. The upright assembly has completed CAD fit checks, with fabrication deferred until that test configuration is needed. Hot-fire testing is pending.

Nozzle-Mounted Pyrotechnic Igniter

Holding position, then clearing the exhaust path

For initial testing, we chose an Estes C-class motor as the ignition source for the engine’s nitrous oxide and isopropyl alcohol propellants. Its comparatively long burn duration was intended to provide an ignition window. I designed the support as a replaceable PLA assembly that could be printed again for subsequent tests. The small ignition motor is single-use; the main liquid engine is intended for reuse.

The main design challenge was giving the support two different functions. It needed to retain the ignition motor against its reaction load, while allowing the assembly to clear the exhaust path when the main engine started.

My initial concept relied on external legs gripping the engine shell, but I was concerned about relying on friction alone for retention. I then considered a plug-style support and rejected it because of the potential to obstruct the exhaust path. A radial shear-pin concept also proved difficult to integrate because I could not find suitable mounting locations.

I ultimately developed an umbrella-style support with folding internal legs. This added parts, but gave me a way to incorporate internal mechanical retention while keeping the custom components straightforward to 3D print. Release during main-engine startup remains an intended function to be evaluated in hot-fire testing.

Refining the physical fit

I checked the design against teammates’ engine models in our shared Onshape workspace before we carried out physical fit checks with printed mating components. The first sleeve fit was too tight, so I increased the clearance; the revised fit was satisfactory. Teammates also reviewed and agreed with the design changes.

We applied weights to the assembled support to examine its retention and failure behavior. In that bench check, it held above the anticipated ignition-motor loading and failed under a higher applied load. This provided a mechanical check ahead of engine testing; numerical test results are not included here.

Upright-Test Propane-Torch Igniter

Integrating protection and actuation

For future upright tests, I designed a separate assembly around an off-the-shelf propane torch. The design needed to accommodate the torch and hose, provide protective structure, and allow the torch trigger to be actuated without making installation and maintenance unnecessarily complicated.

I initially considered separate actuators and levers. As I worked through the layout, the additional parts made assembly and torch installation more complicated. I instead developed a stationary enclosure and moving platform, concentrating the actuation into vertical motion through a lead-screw-driven lifting linkage. The mechanism remains accessible below the enclosure for assembly and maintenance.

I selected 6061 aluminum for the custom parts because it is readily available and straightforward to manufacture. I checked the assembly’s fit in Onshape. Fabrication will require metal machining and is deferred until the team needs upright testing. Torch actuation and thermal protection have not yet been physically evaluated.

What I Learned

This project taught me to think about how a part’s job changes throughout an engine test. Holding the ignition source in place was only one requirement; I also had to consider what should happen to the support once the main engine started. That made me look more carefully at the consequences of each concept, including insufficient retention, an obstructed exhaust path, and difficult assembly.

I also learned that the fewest parts do not always make the most practical design. The folding support added components, but addressed a requirement that my simpler concepts did not handle well. Working through those alternatives, then refining the printed fit, helped me connect the CAD model to the conditions in which the hardware would actually be used.