It's a fun question, Farmflyer. I need to make a post one day which covers more of the behind the scenes steps to bring a model aircraft to life. I'll see what pictures I can get approved to be shown publicly, since yes, most of this is guarded as "secret sauce" type stuff among the manufacturers.
To summarize in a few shorter paragraphs, it's a lot of work!
Selecting a model is based on many factors and data. Designing a model involves drawing it in CAD, then optimizing it for RC and its projected scale. Airfoil, wingloading, stance, servo load, assembly steps, parts count, weight distribution, CG, component layout, etc, etc, etc. The all white models like the one Evelyn is holding is a CNC prototype. Prototypes are CNC'd from dry foam, and tested and adjusted and tested again until the model is flying right. Then, the molds are cut for EPO. Once the molds are cut, there's no going back, because they cost tens of thousands of dollars. The short answer to "how many adjustments do you have to make" is "As few as possible!". In all seriousness, this is where the magic happens. A model could get stuck in a cycle of revisions which takes time, money, and development resources away from other projects. Projects like that either take much longer to release to the public, or they're released prematurely and the public suffers with a poorly optimized model. A well tuned development team can preempt those delays by designing smartly and proactively instead of reactively. After years of work we've gotten to the point that after one or two CNC prototypes of a new model, we're ready to cut the mold. Of course, this isn't a race to just make as many planes as one can. Quality, market saturation, demand, production capacity, and many other factors are considered in the distribution strategy as to when exactly a model is actually released.
Throughout a typical plane's development, prototypes are flown at least 100 times each or more, depending on optimization. I'll fly the EPO version at least 200 times before green lighting mass production. The reason why we fly them so much is to identify wear points, failure points, etc. Data collection is huge. That info is fed back to our Technical Support team's knowledgebase, which in turn benefits the customer when the questions start rolling in. The fun part is long term destructive testing, which includes servo failure simulation where I'll remotely shut off different flight surfaces to see how the airplane flies without them (or if). That's how I know the T-45 can fly with the right aileron completely removed, and the left flap stuck at 10 degrees!!
Where possible, we also try to include feedback from trusted pilots before a model is fully rolled out to the public. This depends on the model, the level of secrecy, logistical and physical access to the prototypes, etc. It cannot always be done, but we do our best where possible. Xplaneguy is one of those trusted pilots, my most trusted non-employee pilot in fact. I've said it before and I'll say it again, I value his opinion not just because we're good friends but because he's straightforward and doesn't hold back if he sees an issue. If he doesn't like something, he says it point blank. And I know he knows what he's talking about after decades of flying. Feedback like that is highly valuable as a developer, because it gives third party insight from a knowledgeable vantage point. Evelyn is highly valued as well, because if she likes something, she makes it known. Her sentiments are close to the pulse of the hobby at large. In fact the Bearcat project was termed Project Evelyn internally, which referenced her personal love of the airplane and her belief, which we share, that the Bearcat was a popular model deserving of positive recognition.
At a certain point once an EPO production sample is available, we do our best to have as many non-Motion RC pilots fly a new model. The P-38 was a good example: Over 40 pilots flew her before the pre-order began. Collecting user feedback is much easier once the models are in the hands of pilots worldwide, of course, but where possible we try to collect that type of information beforehand. Xplaneguy, Bajora, Corsair Nut, JetMang, MiGAceKiller, T-Cat, Air Sally, Mark, Borntoolate, and many others who aren't on any forum (Cory, Rob, Gary, Scott, Isaac, Sam, and Harris come to mind) are just some of the RC minds whose opinions I respect and thus have asked to fly pre-production models in the past. If many of you lived closed, I'd extend the same offer! I urged Crxmanpat to fly the MiG-21 months ago but he shut me down. Hahaha :)
To summarize in a few shorter paragraphs, it's a lot of work!
Selecting a model is based on many factors and data. Designing a model involves drawing it in CAD, then optimizing it for RC and its projected scale. Airfoil, wingloading, stance, servo load, assembly steps, parts count, weight distribution, CG, component layout, etc, etc, etc. The all white models like the one Evelyn is holding is a CNC prototype. Prototypes are CNC'd from dry foam, and tested and adjusted and tested again until the model is flying right. Then, the molds are cut for EPO. Once the molds are cut, there's no going back, because they cost tens of thousands of dollars. The short answer to "how many adjustments do you have to make" is "As few as possible!". In all seriousness, this is where the magic happens. A model could get stuck in a cycle of revisions which takes time, money, and development resources away from other projects. Projects like that either take much longer to release to the public, or they're released prematurely and the public suffers with a poorly optimized model. A well tuned development team can preempt those delays by designing smartly and proactively instead of reactively. After years of work we've gotten to the point that after one or two CNC prototypes of a new model, we're ready to cut the mold. Of course, this isn't a race to just make as many planes as one can. Quality, market saturation, demand, production capacity, and many other factors are considered in the distribution strategy as to when exactly a model is actually released.
Throughout a typical plane's development, prototypes are flown at least 100 times each or more, depending on optimization. I'll fly the EPO version at least 200 times before green lighting mass production. The reason why we fly them so much is to identify wear points, failure points, etc. Data collection is huge. That info is fed back to our Technical Support team's knowledgebase, which in turn benefits the customer when the questions start rolling in. The fun part is long term destructive testing, which includes servo failure simulation where I'll remotely shut off different flight surfaces to see how the airplane flies without them (or if). That's how I know the T-45 can fly with the right aileron completely removed, and the left flap stuck at 10 degrees!!
Where possible, we also try to include feedback from trusted pilots before a model is fully rolled out to the public. This depends on the model, the level of secrecy, logistical and physical access to the prototypes, etc. It cannot always be done, but we do our best where possible. Xplaneguy is one of those trusted pilots, my most trusted non-employee pilot in fact. I've said it before and I'll say it again, I value his opinion not just because we're good friends but because he's straightforward and doesn't hold back if he sees an issue. If he doesn't like something, he says it point blank. And I know he knows what he's talking about after decades of flying. Feedback like that is highly valuable as a developer, because it gives third party insight from a knowledgeable vantage point. Evelyn is highly valued as well, because if she likes something, she makes it known. Her sentiments are close to the pulse of the hobby at large. In fact the Bearcat project was termed Project Evelyn internally, which referenced her personal love of the airplane and her belief, which we share, that the Bearcat was a popular model deserving of positive recognition.
At a certain point once an EPO production sample is available, we do our best to have as many non-Motion RC pilots fly a new model. The P-38 was a good example: Over 40 pilots flew her before the pre-order began. Collecting user feedback is much easier once the models are in the hands of pilots worldwide, of course, but where possible we try to collect that type of information beforehand. Xplaneguy, Bajora, Corsair Nut, JetMang, MiGAceKiller, T-Cat, Air Sally, Mark, Borntoolate, and many others who aren't on any forum (Cory, Rob, Gary, Scott, Isaac, Sam, and Harris come to mind) are just some of the RC minds whose opinions I respect and thus have asked to fly pre-production models in the past. If many of you lived closed, I'd extend the same offer! I urged Crxmanpat to fly the MiG-21 months ago but he shut me down. Hahaha :)
Comment