Full seats + full fuel is generally not the way lightplanes are designed. That’s heavily over-designed for the more typical use case of partial fuel or partial people or partial both. Sure, that reduces the airplane’s utility for the “multi-thousand mile cross country with multiple adult couples on board” use case.
But that’s just the way trade-offs are. You generally can’t maximize everything simultaneously.
Even in big airplanes it’s an issue. Rounding a bit, on my airplane there’s 80,000# of available capacity between empty weight and max takeoff weight. It holds up to 45,000# of fuel, and can accomodate up to 45,000# of people/cargo before we hit max zero fuel weight.
Which means there’s 10,000# that can be people/cargo or fuel, but not both on the same flight. With full payload we have to leave 10,000# of fuel out of the tanks, and with full fuel we have to leave 10,000# of potential people/cargo behind.
Since this is very likely my only opportunity for me to show you something you don’t know about aviation, I’m going to take it - here is a write-up of a fatal crash involving a DC-8 that was primarily due to a series of poor loading practices (but not cargo shifting in flight):
And here is a not dissimilar incident, albeit involving a small aircraft. Particularly pertinent to the current discussion are the references to passenger weights:
Excellent finds both. And those are very nicely written articles. That guy bears reading.
I’d forgotten all the specifics of the Air Midwest accident but I remember when all the assumed weights got bumped in response to a commuter accident.
IMO it’s still the case that the current FAA-approved average weights for passengers (or crew) and their carry-ons are significantly low. On a big airplane the difference is vanishly unlikely to convert “can fly” to “can’t fly” on an otherwise normal day, but for an especially marginal runway and an engine failure at just the wrong time, 2-3 thousand undocumented pounds of weight (i.e. 10-15# per person) could be the difference between passing above some obstacle at the far end of the runway, or through it.
Back to V-tails and a couple of points missed by earlier (excellent) posts:
On the positive side, one of the primary drag reduction mechanisms of V-tails is that there are only two tail surfaces intersecting the fuselage. Intersection drag is a fairly big deal, so reducing it by a third can be a substantial savings. The Beech V-35 was about 5 knots faster than its otherwise identical standard tail sibling.
On the negative side, one of the problems with V-tails is a tendency to ‘dutch roll’, which is the name for coupling yaw and roll. This causes some V-tail aircraft to wag their tails around in even mild turbulence, or while the pilot is inputting roll commands. I’ve been told that sitting in the back seat of a V-35 Bonanza can be nausea inducing.
A V-tail also requires extra hardware to ‘mix’ the surfaces properly on control input. This adds some weight and complexity and cost, but it’s probably not particularly significant in most planes.
If you really want an efficient plane, get one configured with a canard. A conventional tail keeps the plane stable by adding downforce. If the plane goes into a dive, the downforce on the tail increases with airspeed, pulling the airplane back towards level. If it climbs, the airspeed will slow and the downforce will decrease, causing the nose to fall. The same happens in transient situations like wind gusts. If the tail gets pushed down, the angle of attack increases, creating lift (or less downforce), causing it to return to equilibrium.
The problem with this setup is that the downforce from the tail has to be counteracted by the lift of the wing, increasing induced drag. It’s like adding weight to the airplane. But a nose-mounted canard plane maintains stability by having both the wing and the canard create lift, so the airplane flies with less induced drag.
Your other option to reduce drag is to reduce the design downforce on the tail, making the aircraft less stable, or eliminate it entirely, making the airplane unstable without constant elevator input. You can compensate for this with computer control, and some aircraft take this to the extreme such that the airplane is unflyable if the computer quits, as it’s making adjustments many times per second just to keep the pointy side forward…
I have seen some video that some glass cockpits on some fancier GA planes will let you input the weight of each person in a given seat (and the cargo) and, presumably, help you sort out the balance.
If you are flying a small plane do you have to ask the weight of each person coming aboard? Guesstimate? What about cargo (suitcases and the like)? How close to actual values do you need to be before it is a worry? Not really sure if you want to be asking your mother-in-law her weight before a long flight with her right behind you.
Also, can they put a sensor on the landing gear to tell you the overall weight of the plane (since the gear is being compressed by the weight this would seem doable)?
In general terms one of the responsibilities of the pilot and command of any airplane is ensuring that it is loaded within the limits specified in the flight manual.
In a commercial jet the task of calculating weights and balance is delegated to ground staff and the Captain’s responsibility becomes one of ensuring the paperwork is complete and reasonable. In a GA airplane, the PiC should complete a weight and balance calculation for each flight. In small airplanes the numbers of passengers are too low for statistical average weights to work safely, so everything must be weighed, even the mother in law. As an aside, this is also true for large jets if there is reason to believe that the passenger load is not normal. If your A320 or B737 is loaded with rugby teams then they will need to be weighed individually.
That said, it is quite reasonable for a GA pilot to use rules of thumb to ensure the load is correct. Whether it is legal or not depends on the specific wording of the State’s rules, but I have no problem in principle with doing a few generic weight and balance calculations for some edge cases and then ensuring that the loading for each flight is equal to or better than the edge cases.
Any C172 pilot will likely have something in mind such as that they can carry three people with bags or four people without bags and a certain amount of fuel. Even if you do the paper work for every flight, you need to have these figures in mind as a gross error check. For example, if you do a load sheet for a C172 and you find that it says you can take four people, bags, and full tanks, I’d be checking the sums!
For electronic weight and balance for light aircraft, it is reasonably straight forward creating your own using Excel. It is all based on simple sums using weights and moments of passengers, bags, fuel etc. There are also a number of commercial flight planning apps available for tablets that allow you to setup up an electronic load sheet.
Does it matter if the weights are a little bit wrong? Not usually. If you weigh everyone and find you are right on the limit, it is unlikely that anything bad will happen if one of the passengers sneaks a carton of beers on board. Sometimes events conspire against you though and if you are overweight, and it is hot, and the airport is at a high altitude, and maybe the engine is a bit old, etc etc, you might find yourself clipping the trees off the end of the runway. So it probably doesn’t matter, until it does, and you don’t know when that will be, so it is good practice to act as though it does matter all of the time.
Weight sensors are feasible but probably fail the cost/benefit test for light aircraft. I’m not sure where the technology is at for large aircraft, it is certainly something that has been talked about.
I owned a Grumman AA-1B, a two-seat plane with a useful load of 434 lbs with full fuel. If my wife and I were just out flying for fun with no baggage, we didn’t need to do a weight and balance because we were almost 100 lbs under the limit, and with nothing in the baggage compartment. But put two big guys in it, 50 lbs of baggage in the back, and you could easily find yourself over gross weight and with an aft Cg.
Couple that with already poor climb performance and a hot day, and you have a scenario for an accidental stall/spin on takeoff.
This is exactly what happened to Samantha Smith, the young girl who was flying across the U.S. for some political cause I can’t remember. She and her instructor attempted to take off from a high altitude airport on a hot day in an overloaded Cessna 172 with a cargo area full of stuff. The plane climbed sluggishly for a while, then rolled over and crashed killing everyone on board.
You may be conflating the incident with Jessica Dubroff. She was attempting to set the world’s record for The World’s Youngest Pilot (a dubious claim, if you ask me). The accident aircraft was a Cessna 177B Cardinal, and they attempted to take off in poor weather with high density altitude and over gross.
You’re right! It was Jessica Dubroff. I conflated the two.
It was a classic example of being pressured into flying in sub-standard conditions. The press were there, they had a media schedule to keep…
They were flying a Cessna 177 Cardinal, which was already known for its lousy climb performance. It had a laminar flow wing that built up drag fast when at high angles of attack, and laminar flow wings generally have hairier stall and spin characteristics. While the book climb rate was almost as good as a 172’s, the Cardinal’s wing made it very sensitive to optimum climb speed - get behind the power curve and let the angle of attack come up, and in marginal situations the thing could just mush into the ground. Or if it was loaded with an aft cg (easy to do in a Cardinal), you could stall/spin.
Cheyenne airport is at 6,100 ft. On a hot day, density altitudes can approach 10,000 ft. A Cardinal has fairly poor takeoff and climb performance, especially at high density altitudes. I flew into Cheyenne once in a Mooney, and even underloaded by a couple hundred pounds I was careful to not have to takeoff from there in the heat of day. Climbing out of there was a slow process.
The Cardinal is an interesting plane. It was designed to be the modern replacement for the 172. It was faster, handled like a dream, looked way better - and totally failed in the market. Cessna had to re-start 172 production after the Cardinal flopped. At first, it was underpowered, making its climb problems worse. So it got a reputation as a dog, even after it was upgraded with 180hp and a constant speed prop. Too little payload, and too demanding for new pilots who trained in 172’s and wanted to fly what they were comfortable with.
I’ve never flown one, but they’re a hell of a lot better looking than a 172 in my opinion. But after years of flying another plane (the AA1) with lousy climb perfornance, if I ever bought another plane I’d buy one that climbs like crazy. An RV-6 for example.
I iust went back and read the accident report on that. It wasn’t a hot day, but the pilot took off downwind, 96 lbs overgross with an aft cg (within limits, but barely). Conditions were deteriorating, with rain and 1500 ft ceilings. The last pilot to land before them reported ± 15 kts of wind shear. Thunderstorm cells all around.
The plane took off, climbing very slowly. The pilot started turning right to avoid a thunderstorm, then when he attempted to level the wings the airplane stalled and spun.
Cessna said that in the conditions reported and at the weight the plane was at, its theoretical climb was only 387 fpm. But it looks like the pilot didn’t lean the mixture for high density altitude as the book requires, robbing him of maybe 15% power. That may have been enough to put him on the back side of the power curve if he hit a bit of wind sheer, and the 177 is unforgiving if you aren’t climbing at best rate.
Total case of get-there-itis. They should never have flown in those conditions.
The Wikipedia page has an amusing quote relating to that:
We liked that airplane, and found almost nothing to pick at. However, owners soon discovered to their horror that it didn’t fly or land exactly like a strutless Skyhawk, and some heavy-handed Super-Car drivers managed to smash the Cardinal tail into the pavement on landing, knock-off a few nose-wheels, etc. (apparently, this was possible if one closed his eyes, used full back-pressure on the wheel at the flare and then sat rigidly waiting for the crashing noises to subside).
My AA1B had a reputation for being a handful to fly. It wasn’t, but you couldn’t fly it like a Cessna. Instead, you had to fly it like a high-performance plane. You couldn’t just get close to the runway and flare and wait like you could in a 172 - you had to fly it onto the runway just like a Mooney or other high performance plane. And you had to pay close attention to airspeed all the time, because if you let it get behind the power curve it would drop out of the sky like a homesick brick.
I used to fly out of a military field, and they’d let me do ‘overhead break’ approaches (cross center field at about 2500 ft, pull the power, and execute a 270 degree descending turn to the runway). I did that for the same reason the military planes did - it was the only way you could make the runway if the engine quit in the pattern. I was taught to fly patterns such that you could always make the runway if the engine quit, but in an AA1 that was really challenging because it descended so fast with the power off.
The 177 was apparently like that too (without the high descent rate). It looked 172-ish, but with the laminar flow wing it flew like a higher performance aircraft. The people that got into trouble were the ones that had 172 time and figured the Cardinal would fly about the same.
Does anyone here remember how to do Center-of-Gravity computations by hand, or do you all just plug the numbers into your apps to do it now?
Does anyone here remember how to compute that number dear to the hearts of all students, namely, the sacred and feared Grade Point Average?
Has anybody besides me noticed that computing the Center of Gravity and computing your Grade Point Average entail exactly, step-for-step, the same steps? It is exactly the same algorithm. The only difference is the actual numbers that you plug in.
I’m curious about the checks and balances in the procedures. I’m always amazed by large scale processes or engineering projects that rely on large numbers of people trusting each other. What steps are in place to ensure the weighing is correct or not just fat-fingered at some stage? Is there anything? How thoroughly is the loading inspected or supervised? Is there a point in the process where someone could just make a single blatant mistake that isn’t caught? As the pilots do you have much visibility into what people are actually doing? Or do you just trust that everyone else on the ground has done their job correctly?
They’re both just weighted averages–one weighted by, er, weight, and the other weighted by importance. And really, both are equivalent to just straight averages. If you divide your object into equal 1-gram masses, then you can just use a straight average of distance. And if you divide your coursework into equally important points, you can just do a straight average of the grade for each one. The weighted average is just a shortcut for treating them as lumps.
Within the various computer apps there are obviously bounds checks. If somebody fat-fingers a compartment’s load as 100,000 lbs instead of 10,000 that’ll flag if the compartment can only hold 15,000. But if 10,000 is in range and somebody fat fingers 1,000 that won’t. Unless in turn that produces a balance problem. Good bet 1,000 vs 1,100 wouldn’t be caught by anyone. I certainly don’t know all the reasonableness checks that may be included as well.
As much as possible any more, manual input simply isn’t used. A checked bag get a barcode tag generated & attached that’s directly connected to that specific passenger’s ticket records. Then the bag hits an automated scale that copies the weight into the bag’s data record. As it’s going into the can for a wide body or up the conveyor for a narrow body the barcode is scanned by a handheld device and the bag record is individually added to the flight’s load on board. If the barcode is illegible or is tied to a pax on a different flight the tool sets off a siren to stop the loading. Etc., etc., etc.
The same thing goes for the pax. When your boarding pass is scanned at the head of the jetbridge your passenger record is updated to show you’re on board & in the load. It knows if your an adult or child, and where you’re sitting, though as said above it doesn’t know your specific individual weight. It also knows if you’re trying to get on the wrong flight for whatever reason & signals the gate agent to pull you aside & sort this out.
At the end of loading after the doors are closed the system totals the rosters of bags & people. Nobody is keystroking anything.
At our end of the python we're seeing the output, not all the inputs.
But I/we have some rules of thumb. I know how many passengers are expected, and I know how many the final data says we have. I also have a good idea from the FAs of how many we really have. If we’re expecting 150, the FA’s say 2/3rd full, and the final count is 60, I’ll probably catch that it ought to be 160. We also have a pretty good idea of the typical relationship between belly cargo weight & pax count. That worked better pre-COVID and better on the narrow-bodies that don’t carry much dedicated cargo; it’s mostly checked luggage which scales with the pax count. With an adjustment factor for destination and time of year.
Likewise we don’t have the 100% details on the fuel planning. But we do get a breakdown of the various subtotals that go into it and I can catch a gross stupid planning error. And we certainly compare the amount of fuel we’re expected to have loaded vs what the guages show to detect the fueler simply over- or under-fueling us.
It’s a big show and hundreds of skilled people have to do their job right enough often enough. There are enough cross checks stupid shit doesn’t sneak through often. Although as @Dead_Cat’s link to a blog about the Fine Air 101 shitshow clearly demonstrates, there’s lots of ways to screw this up when ill-trained and uncaring workers from 3 different employers are involved in doing the job at 3am.
IANAPilot but I will guess that, as a general matter, you have to trust other people to do their jobs correctly. It is just far too complex for one person (the pilot) to know ALL the things. Pilots have an overview of their plane and how its systems work but they need to rely on maintenance crew to do their job correctly.
I would assume (hope) that all the bibs and bobs of servicing an aircraft to get it flying again have their own built-in checks so there is never one person, if they have a momentary brain fart, that can threaten the aircraft (at least for commercial operations…I would think GA is different since they do not rely so much on others). The pilot gets the OK from the people responsible for various bits and, beyond a cursory check that everything seems to be as it should, assume all is well.
I am not sure how it could be otherwise in a commercial operation.
