Testing the DarkAero 1 Engine Mount

[00:00:00]Hey everyone, this is the engine mount for the Dark Air1 aircraft. It allows us to mount a six-cylinder UL power engine onto the front of the airframe. If you compare it to this example I have here, this [music] is what you'll typically see in a general aviation aircraft.

[00:00:13]Comparing that to the Dark Air1's [music] engine mount, there's several differences. It's clearly a lot shorter in the longitudinal direction and [music] then there's no welding involved in this assembly. This is a really critical structure, so we wanted to test it, make sure it was structurally sound prior to [music] flight testing. Let's get into the design of the engine mount, some of the loads that are imparted into the structure while in flight, how we introduce those loads during the ground [music] test, and then what comes next.

[00:00:39]Let's get into it.

[00:00:43]The aircraft CG is critical to a safe handling, statically stable uh aircraft.

[00:00:49]And with engine mount design, that's kind of where you start with the Dark Arrow 1. We intentionally designed a compact package. And what that allows us to do, if we move the engine closer to the firewall, then we can have a shorter tail or a smaller tail or any version of those variables and ultimately results in a smaller wetted area or smaller aircraft. And with a smaller wetted area, we have less drag. And with less drag, we can go faster, which is what we want. So, this was intentional to get the engine closer to the firewall as opposed to an engine mount like this.

[00:01:24]Uh, and the entire airframe layout is based off of this engine location. So, that's what drives the Dark Air1 engine mount from being shorter in the longitudinal direction. There are some trade-offs here. If you look at our 4130 tube design example, that has its four engine mount locations and then it bridges out or expands out from there.

[00:01:47]The design idea for a mount like that is to stretch out to the perimeter of the firewall and be close to the fuselage skins. Ultimately, the skins is what transfers the engine loads into the aft portion of most aircraft. There are some members, structural members in there, but I'm keeping things high level. So, with it stretching out from the four mount points of the uh engine block, that allows it to have a larger moment arm. And with that larger moment arm, you can reduce the stress at those mounting locations on the firewall uh and reduce the stress in the airframe.

[00:02:22]Uh so with our compact engine mount uh not only being short in longitudinal direction but then being a tighter package in the lateral direction uh we did have to account for uh having a shorter moment arm to react the loads the engine imparts into the structure.

[00:02:39]With that being said, we designed the structure aft of the firewall bulkhead such that we have vertical bulkheads um behind the firewall and brackets uh that allow us to disperse the engine loads into more of the structure uh immediately after the firewall bulkhead uh and ultimately those loads and stresses uh work their way into the fuselage skins and wing and uh rest of the airframe. So there is some trade-offs with our more compact engine mount uh compared to a engine mount that uh extends out and takes up more space within the cowling. Um but our ground test campaign is what we ultimately use to uh verify the structural integrity prior to flight. The firewall bulkhead is a composite sandwich panel material.

[00:03:28]In this bulkhead here, we have a honeycomb core and then it's carbon fiber face sheets that creates a pretty rigid lightweight panel for our firewall bulkhead. And then those vertical bulkheads behind the firewall bulkhead are made from this panel material as well. We cannot just bolt directly to this material. the preload uh of these bolts and then the operational loads that are acting through this region would crush the honeycomb locally where they go through the firewall. So there is some additional design measures and parts locally around where the bolts go through the firewall to make sure we're not crushing the honeycomb. But ultimately the honeycomb panel is what we're using to disperse the loads from the four mount bolts to the rest of the airframe. With this compact design, you can also see that it allows us to rivet on the nose gear mount arms directly to these primary vertical components. Those nose gear mount arms react the nose gear landing loads and then they also house the gear train for actuating the nose gear up and down. One thing to note about this specific prototype engine mount is that it's made of aluminum. We are assessing converting it to steel prior to flight testing. If a firewall forward thermal event was to occur, we could leverage the higher operating temperature capabilities of a steel alloy compared to aluminum. And then the visible carbon fiber on the firewall bulkhead. This is not a flight ready state. As you're looking at it here, we've actually reconfigured the aircraft for Aileron control system proof load testing. So this is a loose installed assembly of the engine mount. Right now for actual flight we would install or will install a fiberfrack and titanium firewall to block off or isolate the composite from the engine and cowling environment. So that's some highle information about the dark air1 engine mount design. Now we could get into the operational loading conditions. The loads that the engine imparts onto these structures can be broken down into four main components. So that's thrust, torque, inertial, and gyroscopic loads.

[00:05:40]Starting out with thrust. Uh that's the job of the engine, right? The Dark Air1 is a tractor configuration aircraft. So the engine's mounted in front of the plane and it's pulling the plane through the air and that thrust is being reacted by the engine mount. and then this uh local structure and depending on where the thrust line of the engine is. So where the prop shaft is located uh vertically in relation to these four mount points uh that location will determine how much or the ratio of that load that goes into the upper bolts versus the lower bolts. That's the thrust. The engine is outputting a torque into the propeller. That's what's rotating the propeller. The air is resisting the propeller from rotating and the aircraft is reacting the output torque of the engine. If you look at a engine supplers's website, you'll see values posted for the output torque capability of the engine or maybe a performance curve. So torque versus RPM, those are good values to have. But with respect to aircraft structural design or engine mount design, that's not the peak load that the engine mount sees. With it being a piston engine, anytime a piston reaches a power stroke, so once the fuel air mixture is combusted and the piston is driving back down to rotate the crankshaft, that drastically increase or spikes the torque output of the engine.

[00:07:01]And so depending on the firing order and number of cylinders that the engine has, uh that creates a different profile specific to every engine where there's pulses in the output torque of the engine. They're called torque pulses.

[00:07:15]And this could increase from the mean torque or average torque. It could increase at 2 and 1 half, three, or even more times that mean torque value. So when designing these structures, you can't look at just the mean torque output from the engine suppliers website. It has to be designed to be able to handle those torque pulses as well where the piston reaches power stroke and there's a a spike in the torque. So that's the thrust and torque.

[00:07:41]Uh and then you have inertial loads. So the weight of the engine when you're flying along in steady level flight conditions, the weight of the engine just as you know if you set it on a scale on the ground, that load or weight is acting on the engine mount and rest of the airframe. And then if you are in a let's say 2G uh pitch up maneuver, then that's twice the weight of the engine that is now acting on um these structures. So as you maneuver and increase the load factor whether that's positive or negative G or in the lateral direction uh you're changing the inertial load that the engine is imparting on the structure. The primary mission of the Dark Air1 is high-speed long range. So this is a normal category aircraft uh meaning it's operating conditions would be uh within plus 3.8 minus 1.52 GS. With our ground test, we did take it past this normal acceptable operating envelope just to make sure we reach proof load values or we have safety factor on top of what the aircraft would normally see in flight.

[00:08:43]So that's thrust, torque, inertial loads. Then you have gyroscopic loads.

[00:08:48]The propeller is a rotating mass and the crankshaft inside of the engine is a rotating mass. So as those rotate depending on how much they weigh and what the mass distribution is about those rotating assemblies and then the RPM of the engine ultimately they have an angular momentum and when the pilot goes to change or maneuver the aircraft change the aircraft's direction wherever the nose is pointing uh that is creating gyroscopic procession with those rotating assemblies. Um so if you start to pitch up there is a gyroscopic load introduced to the engine mount and forward portions of the aircraft uh as a result of those rotating assemblies and that force is acting 90° to the direction of wherever the nose of the aircraft is moving. So those are the four primary loads that are imparted into the structure. So that's thrust, torque, inertial and gyroscopic loads.

[00:09:43]And with respect to design and this ground test campaign, it's important to recognize that those are not independent forces. They all occur simultaneously at the same time on the aircraft. And so with our ground test campaign, we wanted the capability to be able to impart all of those four primary loads into the structure all at once so we can properly study the structure and the integrity of the structure under those superimposed uh loading conditions. We have previously tested portions of this or imparted loads to a certain degree on the engine mount and firewall structure.

[00:10:20]For example, during landing gear drop testing, the nose gear and engine mount and an engine mass simulator were installed on the airframe. So those nose gear landing loads were being reacted through the uh engine mount and into the firewall as well as the inertial loads of that engine mass. So that was a dynamic loading scenario where we have previously loaded the structure up uh before the dedicated primary test campaign uh to validate this structure.

[00:10:48]Uh and then we also did some static loading during the uh horizontal stabilizer and vertical stabilizer proof load tests. During those tests, we used the engine mount and engine mass simulator as well as some additional sandbags to react the loads that we were imparting onto the airframe on the tail.

[00:11:05]We were [music] kind of testing out the engine mount and firewall as a secondary factor of those uh other tests. But then this engine mount firewall bulkhead and supporting structure test campaign was a dedicated test campaign specifically geared for validating the integrity of this structure. If you think about what would happen if the engine fell off or separated from the aircraft in flight, initial thought is that well I would lose thrust. The plane would turn into a glider. Yes, that's true. And hopefully there will be a runway or an improved surface within your glide range that you could land the aircraft on. But more importantly, the engine is one of the heaviest components of the aircraft. And so what that means is if the engine for structural reasons fell off, separated from the aircraft, the center of gravity of the remaining portions of the vehicle would instantaneously shift aft or towards the tail. And that means the aircraft would no longer be statically stable or controllable. So this is clearly a very important structure to validate prior to flight testing. It's pretty interesting. Some race pilots, they'll install a cable or chain between their engine block and then a least likely to fail portion of the engine mount or uh more likely the actual airframe itself. And what that does, yes, you would lose thrust of the aircraft if there was a structural failure. If a engine failure leads to a structural failure, we no longer have thrust, but that cable or chain would hopefully keep the weight of the engine in this region to have a better or higher probability of safely landing the aircraft in the event of a engine mount failure. There's actually a pretty gnarly photo out there from the 1981 Reno Air Races where this happened. In the photo, it's hanging below the front of the fuselage. There's no thrust, but the weight is still attached to the plane, and that pilot was able to safely land the aircraft. I think the engine did contact the ground uh upon touchdown, so it was definitely a rough ride. The aircraft has been reconfigured for Aileron control system proof load testing. But before we reconfigured it, uh we did take some footage of the engine mount test rig uh and test setup to show how we imparted or introduced those four main loads to varying degrees uh into the airframe without damaging other portions of the airframe that we use to ultimately react those loads.

[00:13:32]We've got the vertical load engine mount load test set up here. To start out, we can go over how we are applying longitudinal loads. So, that's through this ratchet strap here. We have a couple straps just to make sure we don't have any unintentional failures while under load. Uh, and then we have a S-type load cell and a turnbuckle. So, as we tension the turnbuckle, we have a readout. Um, and then that tells us how much load we are applying to the engine mass simulator in the forward or longitudinal direction. We made sure to apply this load in the vertical plane at the same position as the thrust line of the actual engine. So if we applied it too high or too low, it would preferentially load the upper engine mount bolts or lower engine mount bolts.

[00:14:20]So we made sure to replicate how the actual engine and propeller is that thrust line in the same location as it actually is in flight. This longitudinal load is reacted with a couple different methods. If we didn't react this load, it would pull the entire airframe forward while we're testing, which wouldn't be good. Everything would be shifting and stuff would quickly get out of control. So, to make sure that the fuselage did not move uh while we were testing, we have these ratchet straps.

[00:14:49]And so these we just have uh lower fidelity instrumentation or or load cells on these just to know what generally how much load we are imparting into this aluminum member. And this aluminum member is constrained to the 8020 frame that's installed inside the cockpit. And that's actually the same 8020 frame that we used for the wing lug load test. And then this is the same method that we used for applying drag load during that test. We're just repurposing things to to stay efficient here. So yeah, these can pull backwards on the aircraft and react that forward longitudinal load. And then in addition to this drag load, we're using the friction of the aircraft. So it's it's setting on top of machined foam. So that's machined to the exact outer mold line of the fuselage and then it's resting on top of cinder blocks. And then to increase the coefficient of friction between the foam and the cinder blocks, we put some sandpaper in there.

[00:15:47]Sticky back sandpaper. That increased friction coefficient between the uh bottom of the foam and cinder blocks gives us ability to react the longitudinal load uh via friction. And then that friction, if we didn't constrain the cinder blocks themselves, they would start sliding as well. So that's ultimately going through this 2x4 frame. uh that you see, we just use spare 2x4s we uh found in the back of the shop and then that's uh going back into this structural pole here. So, um it's just one big loop to react the longitudinal load and then it ultimately keeps the fuselage in the same location while we're testing. So, that's the longitudinal load application method. We could get into the vertical load application method. Now, we're doing that by the same setup essentially as the vertical stabilizer proof load test.

[00:16:39]Uh you may have seen that in one of our previous videos, but just a quick review, that's a hydraulic cylinder that's hooked up to a gantry frame or it's just fixed to the gantry frame. And then we have a couple cylinder extensions here that's compressing this S-type load cell which again has a readout. So as we increase the pressure in the hydraulic system, we can tell how much vertical inertial loading we are imparting onto the engine mass simulator. This position is actually important. So yes, we obviously want it centered between the the two concrete cylinders here, but then we also want to position it forward and aft in a specific location with respect to the firewall. So if we had the actual engine and actual propeller installed, there would be a certain CG location, which we know just how far away from the front of the engine mount that is. During this testing, we want to simulate that same CG position. first found the forward and aft CG of this concrete engine mass simulator. And then we weighed all of the different elements that would be supported by the concrete. So like this HGP frame and then this 8020 frame.

[00:17:45]We'll get into what this is for later, but weighed the different elements, found the static CG location without imparting any load onto the assembly.

[00:17:55]And then we know how much our target proof load value is for the vertical component of this test. And then you can combine those together to find a resulting effective CG position while we're at peak load. So we want that effective or simulated CG position to match the real engine and propeller CG positions. These aren't just placed here randomly on the concrete. It's actually measured out and then double verified exactly where we are placing these forward aft. similar to how it matters where this longitudinal load application method is in the vertical plane or water line of the aircraft. The vertical load is imparted through this hydraulic cylinder here and then we're reacting that vertical load through all of this counterbalance. If we didn't then the aircraft would pitch nose down as we increase the load. So that's why you see uh concrete here in the cockpit. This is where the occupants sit and we have sandbags in the a bagage compartment.

[00:18:53]Additional sandbags along the empanage here and then more on the horizontal stabilizer. This is kind of a horizontal stabilizer static load test in its own right, but for the actual horizontal proof load test, uh we had the load dispersed all the way out to the uh tips of the horizontal stabilizer. So, we're just centralizing or bringing the load in uh towards the root of the horizontal stabilizer so that we're imparting less risk onto that assembly. uh when we're trying to intentionally load up the front of the plane. So that's the vertical load application method. We can get into uh what this is doing now. So the engine creates a torque on the rest of the aircraft and then we're imparting that torque just with repurposing these large pieces of 8020 framing. This extends down through between uh these cement cylinders. And then uh we are pulling on that and reacting that against the gantry frame. Uh there's a lower fidelity load cell up there and then a turnbuckle and a ratchet strap.

[00:19:56]So as we turn the turnbuckle, we can increase the tension and then we know the distance from that lateral load acting point. We know the moment arm down to the rest of the engine mass simulator and then that gives us a torque to simulate the engine torque. So with this torque application method, it's not actually a pure torque. So, we're applying a lateral load at a distance from the center point of our engine mass simulator, which means yes, we do have a torque or a moment on the assembly, but we also have a lateral load that we need to react. So, that lateral load, we are reacting via this strap. So, that's running over in towards the co-pilot side of the aircraft. And then if we apply say 10 lb in this torque inducing turnbuckle strap, we can apply 10 lb in this lateral strap here. So that at that point we have a pure torque moment on the engine mass simulator and no lateral load component. This does produce a gyroscopic moment on the engine mass simulator which is intentional. We do want to impart gyroscopic moments on the assembly. So this allows us to impart some amount of gyroscopic moment. And then if we want to impart even more uh in addition to a lateral load, we can increase the tension on this strap down here. One thing I forgot to mention here is what all these concrete bags are doing. Just the sheer magnitude of vertical load that we're imparting on the front of the aircraft here requires us to weigh down the gantry frame more than its out of the box weight. So, this is allowing us to apply more load uh in the vertical direction than the weight of the gantry crane. We could have bolted this into the shop floor, but this is easier to keep things mobile.

[00:21:44]What this strap out here is doing is reacting the torque load. So, as we apply load via this strap to that A20 frame, if we have the aircraft constrained so that the aircraft doesn't move and we did not constrain the frame, then the frame would just start rolling that way. So to react that we have this strap. It's going down and secured into the floor so that the gantry frame uh does not move on us. So that's how we are imparting torque into the engine mount and then reacting it. So if we do apply lateral load to the engine mass simulator, we are reacting that component of the load with the barrels.

[00:22:24]As you see, if we pull on the front of the aircraft that way and we don't constrain the aircraft in that direction, then the aircraft would yaw to the right. So, um, there's a barrel there. It's just filled with grit blast material. It's essentially just a heavy object we're using to counteract or resist the airframe from moving. And then that's pressed up against the side of the aircraft with a piece of foam to make sure we're not damaging the sidewall of the fuselage. That would not be enough. If we just had that barrel, it would act as a pivot point and the tail of the aircraft would still kick out. That's what this barrel over here is for. With the same method, we've got foam up against the side of the uh empanage/vertical stabilizer. Uh that's preventing the tail from kicking out. So uh this is how we are reacting the lateral loads we're applying to the front of the aircraft. We could again build a structure and bolt it into the floor of the shop, but just keeping everything on wheels and using the natural weight and friction of things allows us to stay mobile and constantly reconfigure our test setups if things aren't right or if we want to modify things. We don't have to continually move anchors in the shop floor or reconfigure a welded jig or something like that. A lot of these load application methods are using turnbuckles to apply the load or we have the hydraulic cylinder. And all these load application methods are safe from a failure standpoint because as soon as a failure occurs and the distance between the two points of the turnbuckle shortens, all the energy will safely leave the system. We won't have too much elastic energy whipping things around.

[00:24:01]So, keeps us safe and the airframe safe.

[00:24:03]I think that's majority of aspects I want to go over for the vertical engine mount load test setup. So, we're going to get into testing and then eventually reconfigure it for negative uh glo engine mount testing. In addition to the testing configuration uh where the aircraft was in the upright orientation or inverted uh we also tested the aircraft while it was on its side. So laterally loading the engine mount and firewall bulkhead structure and that is for any yaw accelerations whether intended or unintentional from things like gusts. Overall the test campaign went fairly well. We were able to find a weakness in the upper engine mount bolt supporting structure behind the firewall but because we were able to increase and decrease the loads during test in a very controlled manner. we were able to prevent any catastrophic or significant damage from occurring to the aircraft and then we uh installed newly designed brackets behind the firewall and once those were installed then we uh proceeded forward with testing until we had passed all of the predetermined test cases and proof load values that we were looking to pass prior to flight testing.

[00:25:13]So the engine mount firewall bulkhead and supporting structures are now structurally validated for flight testing. So that's pretty exciting for the team. If any of this sounds interesting to you, whether it's loads definition for structures like these, designing the structures themselves or building flight hardware or designing test rigs and testing flight hardware, if any of that sounds interesting, you should check out our careers page on our website and apply to join the team. I think this is sick. Okay, with the engine mount load test campaign wrapped up, we're moving on to ground testing the control system and we're starting out with proof loading the ailerons.

[00:25:48]That'll have to wait for another video, though. We'll catch you next time.

[00:25:54][music]