With the insights gained from Tiger III, it was time to develop Tiger IV, the first rocket designed to attempt an actively stabilized flight.

Tiger IV is entirely 3D-printed using PLA and features my first flight computer, Lion I.

CAD

Tiger IV features an internal rigid structure with a screwed-on casing to enhance its aerodynamic profile.

It also includes very small ascent fins, designed to help position the center of pressure close to the center of mass during normal-speed flight operations. This arrangement makes it easier for the TVC system to control the vehicle effectively.

TVC

The TVC system on Tiger IV features a unique design. The engine cluster is mounted on a spherical hinge, allowing it to rotate freely in all three axes. Two pistons, each attached to the engine mount with additional hinges, adjust the thrust angle by extending and retracting. Instead of pneumatic or hydraulic cylinders, the system uses servo motors. Each servo’s rotational movement is converted into linear motion with the help of an extra hinge on each piston rod.

Given that servo movement is not entirely linear, the conversion between the servo angle range and the TVC angle is complex. To simplify this, I developed a conversion table for the flight computer to determine the appropriate servo angles.

While the TVC system provides a very fast response time, it encounters challenges with play in the three hinges. Since all hinges are 3D printed, they must strike a balance between minimizing play and managing friction.

Parachute Deployment System

Given that the parachute deployment system of Tiger III didn’t work as anticipated, I had to devise a new concept for Tiger IV.

In total, I went through three completely different design iterations to find one that would be sufficient for the application.

Spring Loaded

The first method involved a spring-loaded mechanism. The nose cone is pushed away from the parachute hull by four springs positioned around its perimeter. A rubber band holds the nose cone in place. Below the parachute hull, a servo with a point-symmetric horn is used to remove the rubber band. Once the band is removed, the springs would detach the nose cone, allowing the parachute to deploy.

The compression of the springs proved insufficient for effectively propelling the nose cone away from the parachute compartment. Consequently, I moved on to the next design concept:

Explosive

This parachute system was one of the simplest I’ve ever designed, utilizing explosives for deployment. The parachute compartment in the center is surrounded by a ring of black powder. The nose cone, positioned on this ring, features a surface designed to direct the explosion's gas to push it away from the compartment.

However, this system did not work as intended. The black powder used was of insufficient quality, failing to generate the necessary amount of gas for proper deployment.

Side Deployment

The final iteration uses a different approach. The parachute hull features a side opening, and inside the parachute compartment, a cut-open plastic bottle acts as a spring to propel the parachute out. The side opening is secured by a door attached with a rubber band that wraps around the parachute hull. One end of the rubber band is held by a servo motor. When the servo motor rotates, it releases the rubber band, unlocking the door and allowing the bottle spring to push the parachute out.

This concept proved successful in multiple tests. You can see it in action in the video below:

Launchpad

The launchpad is constructed from wooden plates that are doweled together, featuring a right-angle wastewater pipe as a flame trench.

It is powered by a lead-acid accumulator capable of providing high currents. The system uses two Arduino Uno R3 boards to provide the necessary GPIO pins and processing power.

There are six indicator LEDs that show the status of the pyro channel, the regulator, and the battery power.

On the left side of the launchpad, three switches are available: one for powering on the system, one for arming the pyro channels, and one for initiating the countdown sequence.

During the countdown, the launchpad’s RGB LED strip and speaker create a countdown animation. At T-0, the pyro channel ignites the four engines of Tiger IV using a MOSFET designed to handle high currents.

The launchpad communicates with the Tiger IV rocket via a three-wire serial cable, which is detached by the force generated during launch.

Here’s a video of the countdown sequence and ignitor activation:

Lion I

Features

Lion I is the first PCB I ever developed. It is powered by a standard 9V block connected to a screw terminal on the board.

The 9V input is step-down regulated to 5V using a 7805 regulator.

The board features a Teensy 3.1 as the processing unit. It also includes an SD card breakout module for storing flight data, as well as a BMP280 breakout board for assessing altitude (barometer), and an MPU6050 breakout board for determining orientation with its 3DOF gyroscope. Both sensors have their own 3.3V regulators, allowing them to be directly powered by the 5V supply.

For output, the board provides two servo ports, an RGB LED, an active buzzer, and two pyro channels.

Problems

Even though the board was functional, it had several design issues:

Firstly, I encountered problems with the BMP280 and the pyro channel screw terminals. I accidentally chose the wrong pin hole diameter for the BMP280, making it impossible to install the component properly. Similarly, the screw terminals for the pyro channels had incorrectly sized holes, requiring me to drill them out to accommodate the components.

Another significant issue was with the DIP switch. It was positioned directly in the power delivery line but was only rated for milliamps, while the battery power line can carry several amps. This mismatch posed a risk of failure due to the high current requirements.

Additionally, the 7805 regulator, which supplies 5V from the 9V input, was insufficient to power the two servos. It can provide up to 1A, but each servo can draw around 1A on its own, leading to potential power supply problems.

Launches

In total, the rocket was launched twice, both times using C2-P engines with a burn duration of 5 seconds and a thrust of 2N.

Unfortunately, on both launches, the engine ignition mechanism failed. Each time, one of the four engines exploded instead of propelling the rocket.

Initially, I suspected the cold weather was the cause of the failure. However, upon further consideration, it appears more likely that the issue was due to the electric ignitor being inserted too tightly. This likely caused the engine to clog, preventing it from releasing energy through the nozzle as intended.

In both scenarios, the rocket began to spin heavily immediately upon liftoff. The footage does not clearly show whether the TVC system attempted to counteract this rotation. This intense spin is likely due to the explosive force from the engine failures. Additionally, the right-angle flame trench might have contributed to an uneven recoil force, further complicating the situation.

The rocket was a significant improvement over Tiger III, but it still wasn't developed enough to achieve an actively stabilized flight. The TVC system had too much play, the avionics system was error-prone, and the deployment system had issues.

As a result, I began incorporating these learnings into the next line of rockets, which would be known as Buffalo.

Project by Johannes Moser

Tiger IV

CAD

TVC