I know these things are going to break

Picture of Adam Bahret
Adam Bahret
Screen Shot 2022 10 11 at 1.48.30 PM

So the Alphabet engine is coming along. Although I want to get to the completed running motor quickly, I want to do things right so I can enjoy this system day in and day out for many years. I don’t just want to have some rushed prototype that eventually ends up on the shelf because it turns out to be problematic for day-to-day use. The fuel rails I have currently are made of plastic. As their name suggests, these are the components that bring the fuel to the fuel injectors. In the system, I currently have ones from the MV Agusta motorcycle that the throttle bodies came from.

I kept the spacing between the fuel injectors the same in my setup as the bike, so why not use a ready-made fuel rail? Well, here is why I recently decided not to. Even though these stock rails are well-proven in the field, they may not be equally as robust in my application. Usecase!

Let me reiterate that these rails are plastic, not metal. Today, this is common in most car and motorcycle applications. Plastics have come a long way and can be trusted to handle high fuel pressure for long-term applications without issue. They are cheap to manufacture and easy to quality control. Moreso, they are often replacing what would be several metal parts welded together with only a single plastic-injected molded piece. This translates to big cost savings.

So, if they are so robust, why am I concerned about them in my application? My application, although similar, is not the same. In car and motorcycle applications, the rails are put in locations that keep them well protected from being nudged, or exposed to UV, water, or even harsh language.

However, and here lies the problem, that is not the case in my application. They are right out in the open, even exposed to rain and light (the Porsche has a grill right over the motor). When I fix and tweak things, I will have to reach right past them in order to get to just about everything, I might even lean on them as I fuss to get some other part out. So in this application they will be nudged, bumped, exposed to light + moisture + heat (the holy trinity of plastic degradation), and surely harsh language.

What I am doing is considering preventive maintenance in my first design iteration. In this case, it’s leading me to increase robustness, not create another PM procedure. Why simply design for more robustness? It will be very hard to determine degradation from the bumps and nudges. It is a highly variable use case stress that would require extensive field studies with many scenarios. None the less in this application I fear there could be enough damage for a high-pressure leak from even a single incident. It would take a very extensive and conclusive study to convince me otherwise. But considering PM brought me to realize this usecase is very different.

It is very difficult to effectively design for service once a design gets to its later stages. I am confident you all can attest to this, but so often forgotten until we are ready to transition our prototypes to final production. So for my application, we designed and built a set of aluminum fuel rails. We added some other features like integrated pressure dampers and improved connection types while the drawings were back on the table.

Now the big question is, In new product development programs how early in the design process should we think about preventative maintenance (PM)? I think the short answer is to keep people who are knowledgeable about PM in the know starting at the program kick-off. PM cycles are created based on performance degradation studies. Degradations in performance, part health, and risk to instant failure are all considered. Once these maintenance schedules and procedures are determined, they are optimized to keep total product downtime at a minimum. Remember the uptime we advertise to our customers includes scheduled downtime (PM) as well as failures. It doesn’t matter why it is down, the system is down and not serving the customer.

I so often see teams not consider the stress of PMs, like in this case. Not just to the serviced item but stress from service to nearby items as well. A common example is what can happen to wire harnesses. PM instructions can say, “Unplug harness “A” and move it clear from the working area” Was that plug on harness “A” only spec’d for one insert/removal? Or were the 20 cycles over its life from the PMs, direct or peripheral included in its use case?

Plus, how about the other end of the harness we are not unplugging, the one that plugs right into the PC board? The receiving connector soldered to the board is getting cycle stresses when the harness is moved. Are those eight soldered pins going to hold up under that stress? Or maybe worse it is firmly fixed to the PCB board to ensure the pins aren’t stressed. The result is that the cycling stress now transfers to the board itself. PC boards have little patience for being asked to do diving board impressions.

Final thoughts: you should start bringing up best practices for PM in design reviews early and I’ll keep considering what I can do in the Alphabet engine to minimize the chance it gets on the 7:00 News as a local fire story.


Was this article forwarded to you? You can sign up and have them dropped right in your inbox. Here is the super slick sign-up page.


Share this post