Launch vehicles powered by rocket motors accelerate rapidly and their guidance systems are extremely complex. Imagine maintaining control of your car with the cruise control set to 17,000 mph. “We used the classic approach in designing the wiring harness for the Ares I booster,” said Nathan Holyoak, Engineer PLM Processes for ATK. The Ares I is shown in Figure 1. Technicians built a full-size mock-up of the Ares first stage integration components and ran rope through it to simulate cable runs as shown in Figure 2. Wooden boxes were used to simulate electronic components. Different diameters of rope helped determine how many wires would fit in a given area. The cable information was then captured in a massive spreadsheet that was used to store information on each cable and perform a variety of calculations such as voltage drop. Electrical schematics were created using Microsoft Visio and mechanical computer aided design (MCAD) drawings were created using the Siemens NX CAD system with a sweep along a spline with a fixed diameter (often referred to as space reservation). ATK also built a testing lab using engineering workstations to perform a wide range of simulations and analyses on the wiring harnesses. After the review, the prototypes were developed.
“It took several full-time engineers to maintain the mock-up and the spreadsheet was so large and complex that managing it was very resource-intensive,” Holyoak said. “It was difficult to ensure that the spreadsheet entries were correct and up to date, so additional time was spent measuring and performing calculations over and over again. Our MCAD drawings showed nice pictures of harnesses but told us nothing about their functionality. Generating schematics and drawings with Viso was a long, drawn-out manual process and changes were time consuming. In addition to the less effective prototype method, the lack of accurate electronic formboard documentation meant the cost of producing the wire harnesses was excessive. Additionally, the prototype’s physical location was not central relative to all the design teams, which added extra travel requirements.”
NASA changes programs
ATK and NASA wanted to leverage work and experience from the Ares I for the SLS, yet the SLS design was different enough from Ares that the wire harnesses needed to be essentially designed from scratch. Under the terms of the SLS booster contract, ATK was tasked with testing and delivering the new boosters with a team that was significantly smaller in size than the one that developed the harnesses for the Ares I.
“When NASA came to us and said they wanted to change direction, it was an eye-opener to discover that we had made a significant investment that suddenly had little to no value,” Holyoak said. “Despite the fact that the SLS boosters were based on the Ares design, the differences were large enough that they would have required starting from scratch by constructing a new prototype, new spreadsheet, and new drawings. Very little from the previous project was reusable. We were also conscious that our previous methods were very labor-intensive, so when we discovered that our next project would be equal in magnitude yet would have to be done with less manpower, we realized we needed to improve our process – and do it quickly.”
Developing a digital approach
“We decided to develop a digital approach that was less labor-intensive and would enable us to more easily adapt to change,” Holyoak said. We wrote a set of requirements for an electrical CAD (ECAD) system that would be both logical and functional. It had to communicate with our MCAD system so we could quickly determine whether a cable would fit or not and how long it needed to be without mocking it up in a lab. We also insisted that the ECAD system be integrated with the MCAD system and that the results be controlled so that only authorized users could make changes and the change history would be recorded. We selected E3.series because it met all our requirements.”
Examples of typical MCAD and ECAD drawings are shown in Figures 3, 4 and 5.
ATK worked with experts from Zuken and Siemens PLM to get the two software packages working together using Zuken E3.3D Routing Bridge, which communicates using PLXML. NX librarians and checkers created over 3,000 new or modified piece parts to ensure the libraries were synchronized. Outputs from Zuken’s E3.series and Siemens’ NX are released and controlled through Siemens Teamcenter, which also handles released engineering data.
Each placement of a component in the E3.series project provided intelligence that automated many of the downstream activities that in the past had to be performed manually. E3.series ensured consistency of the entire project by keeping track of the logical design and ensuring that the physical design was consistent. The software checked the consistency of connectors, ensuring, for example, that connectors mated with each other and there were enough pins to handle each wire in an assembly. E3.series automatically tracked the components used in the project.Siemens NX calculated wire lengths and passed the information back to E3.series, which in turn calculated the voltage drop in the cables. Instead of manually producing drawings as required in the past, engineers were able to generate alternate views of the wire harnesses for formboard or cable documentation as required with very little effort.
Fewer people working smarter
Figure 2: Mock-up of Ares I booster
“In the past we had a lot of people working hard,” Holyoak said. “Today we have a much smaller number of people working smart, with an automated interface, and the results are impressive. The electrical engineers enter requirements and the software makes sure they have been met. If there is a problem, E3.series greatly reduces the checking time by enabling the engineer to simply click on a signal to trace its route through the design. With a virtual logical and physical model, the teams of engineers save time by not having to travel to get to the prototype lab. Instead they can just view the design on their computer screen. With the automatic checks performed by the software, people have a much higher level of confidence that everything is right, so we are saving a huge amount of time that was previously spent checking and double-checking. We talked to our suppliers and asked them how much they would charge us to produce the formboards themselves if we provided them with the electronic documentation. We have to include some extra information in the drawings, such as braid overlaps, but compared to the original cost, the savings were substantial.”
Figure 3: Typical MCAD drawing in Siemens NX
The cables for the entire SLS program were developed in E3.series and Siemens NX without a physical model, so the prototype lab was closed. With verification automatically performed within E3.series, the test lab was no longer needed so it was repurposed into a verification lab. The SLS project, which is similar in magnitude to the Ares I project, was completed in roughly the same amount of time with significantly fewer resources.
“Now we are getting requests for different configurations of these boosters,” Holyoak said. “In the past it might have taken us months to design the wire harnesses and agree to a POC (proof of concept). Now we can do a feasibility test in weeks by copying and making changes to the existing model. Going from months to weeks has not only made ATK more competitive, but has established us as a trusted partner of NASA. Besides these savings, we also have much more confidence that our designs are correct. Zuken has been an exceptional partner throughout this process and was there with us the entire way.”
Figure 4: A typical MCAD drawing in Siemens NX
Figure 5: Typical ECAD drawing in Zuken’s E3.series