Causes of Engine Complexity
Purpose of Article II
In this article, the reasons for Turbo-pump complexity is discussed. We look specifically into inert gas seals in Turbo-pumps, and how they impact the number of parts and the cost of the Turbo-pump (adding 200 parts). This gives a good idea on why and how offering an alternative to shaft seals can lead to a significant complexity reduction.
In the article named “Money Pits in Rocket Production”, it is justified (and quantified) that the physical complexity of an engine has a lot of influence on the cost of a rocket. They way it can be avoided is by decreasing the number of components inside of the engine in order to make it more simple.
One of the causes of the physical complexities of a Turbo-pump is the shaft seal mechanism that they utilize. The nature (and functioning) of the seal brings with itself a large amount of small parts, increasing the cost.
More Context Into Shaft Seals
Before getting into the design of the seal, it’s important to understand why the seal exists in the first place. The function of a Turbo-pump is to pump oxidizer/fuel at really high speeds into the reaction chamber by applying rotational kinetic energy to the main turbine which accelerates the oxidizer/fuel.
As engineers have designed the Turbo-pump to somewhat fit the chart above, they have come up with a design that is space-efficient, and also reliable. In short, because of this design, Turbo-pumps handle oxidizers and fuel really close by.
In simple terms, a Turbo-pump functions as the following:
- The purple liquid (coming from the pre burner) propels the turbine (grey) due to its forceful flow.
- The purple liquid enters in through the very right, interacts with the shaft before exiting from the points labeled “O”
- As the turbine spins faster because of the purple liquid, it pushes forward the red liquid which enters through the points labeled E, and exits with a higher rate of flow through the very left.
- As these fluids are interacting, high-pressure helium (blue) enters through the parts labeled H to prevent the reaction of the red and purple liquids. This is the general idea of purge seals.
Diving Deeper into Shaft Seals
There are not any great numbers out on the web pointing out the number of parts involved with the Shaft seal, so to quantify this concept of complexity, an educated estimate will have to do for now.
Using diagrams from various different sources, we can figure out that the Turbo-pump shaft seal is a series of seals each reducing the gas leakage rather than a single seal. This series of seals includes:
- Wear Ring Seal (x1)
- Floating Ring Seal (x3)
- Pressure Labyrinth (x3)
- Segmented Seal Rings (x2)
Along with these seals, the system needs:
- Pipes to deliver the inter gas (3 part pipe)
- Valves to ensure one-way flow through pipes (x2)
- Housing for the series of seals (metal parts that keep the seals attached to the casing of the Turbo-pump)
Based on diagrams published by manufacturers & designers, we can approximately count the number of individual parts that exist within each of the listed components above.
- Wear Ring Seal — 1 Part
- Floating Ring Seal — 4 Parts (x3)
- Pressure Labyrinth — 2 Parts (x3)
- Segmented Seal Rings — 25 Parts (x2)
- Pipes — 3 Parts
- Valves — 50 Parts (x2)
- Seal Housing — 50 Parts
In total, this brings us to ~200 parts, which all exist to make the inert gas seal.
Along with the fact that each of these 200 parts have to individually traverse the supply chain, it is also important to keep in mind that each of these seals require a lot of testing before being driven into production.
The reason for this is that the status-quo of seals are physically attached to the spinning shaft in the middle of the turbo-pump, but also attached to the casing. This subjects it to a bunch of pressure and load, giving it more data-points to be tested on. Some of the tested variables involved are pump pressure, pump flow rate, shaft rotational speed, turbine pressure, turbine gas flow rate, shaft power, weight, shaft power etc. This isn’t to say that the proposed alternative will not be tested, but rather point out that the amount of testing will be dramatically decreased along with the number of parts. With less moving components and minimal interaction with the spinning shaft itself, the proposal (explained in article III) will be much easier to experiment upon.
It is difficult to give an exact amount of time, and money spent on testing these seals, but a report published by NASA, and a design criteria overview paper puts the amount of testing required into good perspective. In the end, we know that testing in Turbo-pumps do take up 40% of their production costs. This is not as surprising when faced with the thousands of testing reports presented in these articles.
Conclusion
80% of the cost of creating a Turbo-pump is directly influenced by the number of parts due to testing and supply chain expenses. To drive down the cost of Turbo-pumps, we realize that it is important to identify a point of improvement, which can significantly reduce the number of parts, therefore pulling down the cost. One way to do this is by coming up with an alternative to Inert Gas Seals, used by traditional Turbo-pumps. There exists 200 parts to keep this seal working, that can all potentially be eliminated with a simple alternative.
