Menu Sign In Contact FAQ
Banner
Welcome to our forums

How many hours to build a GA piston aircraft?

It's a matter of production volume. Volkswagen are aiming for 10 million cars a year, that is 27400 cars a day. The civil aircraft with the largest volume in history is the Cessna 172 of which ca 43 000 were built since 1956. Volkswagen builds more in less than two hours. Nowadays, Volkswagen builds more than twice the number of cars per hour that Cessna builds airplanes in a year.

So you just can't afford the serial production technology like robots and conveyor belts. If you try to get there, you have to fool investors and the rest of the world (Eclipse).

And your example A2 is a good one because it shows that even car makers with their huge volume can't afford non standard manufacturing techniques because the market does not honor innovation that increases the price.

I still think that much could be done in GA manufacturing.

For example nobody should be wiring up the main cable harness in situ. These should be made up, including every conceivable option, on an external jig.

Socata went a long way to improve things, with loads of jigs and press tools, but even they had loads of manual labour.

The car industry process whereby you press out a whole panel, with composite curves, will never be possible for GA because the tools are so expensive. Socata produce some of that kind of stuff (e.g. TBM cowlings) by stretching the material over a mould, but even that is a massive piece of kit with a throughput of about 1 per hour (I've seen it operating). The GA business seems to have a choice between (a) non-composite curves, and (b) making composite curves from composites (no pun intended) which are labour-intensive to lay up and generally not any lighter (but they look great).

But then this is all hypothetical because almost nobody is making new planes nowadays

Socata packed up pistons 10 years ago, and make the TBM using 1950s methods. Same with the PC12, and most bizjets.

That leaves Cirrus which are composite and they are the main player. All the other stuff (Beech, Cessna, etc) are basically 1950s designs and build technology.

We're torn between state-of-the-art glass panel and 100% basic, authentic, traditional, analog. We've seen both and can't decide which to go for.

How much value would you attach to the ability to replace individual instruments, possibly a long way away from base? If you always fly from a specific farm strip to several nearby locations (burger runs) then go for the integrated kit.

However I know nothing about the quality or reliability of the integrated kit in that market.

Administrator
Shoreham EGKA, United Kingdom

I'm not sure I agree with you 100% on the A2 technology. The A2 was discontinued due to poor sales, only 175000 or so (iirc), but the technology lives on in the A8 (which came first in fact), the TT, R8 and other models. They all use a great deal of aluminium and A2 derived processes.

I can't see how the work hours of an armada of rivet gun operators could be more efficient than a riveting robot, or welding robot. Yes, it's a matter of scale and we'll never get close to an auto manufacturers weekly production run (not the big giants anyway), but if you look at the amount of labour that goes into a Cirrus for example it boggles the mind. The manufacturing process is so inefficient they struggle to make a profit even on an $800k airplane. Why bother?

Did anyone see the Disney cartoon where D.Duck bakes his own plane in the oven? That'd be something...

ESSB, Stockholm Bromma

I can't see how the work hours of an armada of rivet gun operators could be more efficient than a riveting robot, or welding robot. Yes, it's a matter of scale and we'll never get close to an auto manufacturers weekly production run

Why do carpenters in Egypt have hand drills and saws instead of drilling machines and buzz saws? They could greatly increase their output, margin, profit. Because it requires capital and someone to make that capital available and the willingness to increase the risk by taking higher loans.

Eclipse showed it: billions of dollars spent before the first plane got built and the moment the optimistic plans didn't work out, it all collapsed. Cirrus started by building experimental airplanes and steadily improving them. They made a big bet with their jet and it almost got them bankrupted.

It's pretty clear that Cirrus wouldn't sell a lot more if the airplane costed $600k instead of $800k.

It's pretty clear that Cirrus wouldn't sell a lot more if the airplane costed $600k instead of $800k.

That's probably true, but only because they have no competition anymore.

Amazing when you think my TB20GT, loaded with just about every factory option possible (short of aircon) cost $300k in 2002...

Administrator
Shoreham EGKA, United Kingdom

Hmm, you're not seriously comparing egyptian carpenters to american rivet gun operators are you? We might as well compare apples to oranges.

The idea taken from the A2 would be to stamp out a few preformed parts that weld together instead of cutting 5000 individual pieces of Dural bent over wooden blocks and assembled with 10000 rivets.

On the other hand, what should an aircraft cost? Are we talking €20-30k or €200-300k? Would sales increase that dramatically if there were cheap planes? In fact, there are relatively cheap planes, or microlights/ultralights/VLA type, but they're still not selling by the 10 x thousands.

Perhaps the issue of aircraft is not their cost, but the complexity of learning to fly, maintaining a license, owning the plane, making use of it?

ESSB, Stockholm Bromma

How much value would you attach to the ability to replace individual instruments, possibly a long way away from base? If you always fly from a specific farm strip to several nearby locations (burger runs) then go for the integrated kit.

Peter I think this is far less true these days. Firstly the equipment is far more reliable. But even more significant is the uncertified avionics are truly cheap for what you get. I would go for that - but then I would say that wouldn't I?

EGTK Oxford

Perhaps the issue of aircraft is not their cost, but the complexity of learning to fly, maintaining a license, owning the plane, making use of it?

Yes I think that is a really big limiting factor on GA growth. To get any real A-to-B capability you need an IR... that's a "life project".

With all this said, I still cannot see how the costings for say an SR22 (taking the one significant plane in current production) add up.

Taking the TB20:

The airframe metal comes to roughly €1000. Not kidding! That is the cost of the aviation grade aluminium, delivered to Socata with the release paperwork. Then you have various bought-in custom-made parts e.g. the elevator jackscrew mechanism - my guess is that these come to another €1000, and these don't have to come with release paperwork because Socata is EASA145 and they can inspect them in-house.

Then you have the bought-in parts e.g. the engine, avionics, door locks, etc whose cost has to be guessed at. An IO540-C4 lists at about $60k so I guess an OEM like Socata might be paying say $30k. The avionics won't be cheap - probably another $50k including the autopilot.

So we are looking at about $100k's worth of parts, max.

They used to sell it to a dealer like Air Touring at list - 15% i.e. $255k.

So we have a gross profit of about $150k or 60% of the selling price.

That's actually pretty good. A "specialist" manufacturer would normally work on at least doubling the material cost to get the lowest (reseller) selling price. One tries to go for 2.5 times. So this looks about right.

Don't let anybody tell you bollox about not making a profit in aviation

The problem, obviously, is whether you make a net profit, after throwing in the other costs.

To be fair, before doing the 2x multiple, one should bring in the variable costs i.e. the direct labour. This will be quite a lot. In 2002 it cost about €80k to employ 1 relatively skilled worker in France (their employers' contributions are about 50% of the gross salary!). You have about 1500 man-hours in a year. Let's say a TB20 takes 500 hours of direct labour - that comes to €25k. Not really a big dent...

So the real money sink must be elsewhere.

It can't be the certification, because every plane we look at is either very old (and all the cert was done decades ago) or is less than very old but the ongoing cert work is in small incremental improvements. The FAA gets involved only in bigger stuff like an MTOW increase.

It leaves us with the fixed costs of the establishment, as the most likely black hole.

If somebody inside the business can put better meat on the bones, please do so (actual figures I mean).

Peter I think this is far less true these days. Firstly the equipment is far more reliable.

I am absolutely certain it is not more reliable, in the general case, if you ignore known junk like Narco.

There is a pretty good chance you will lose your G500, and much more likely you will lose an Aspen, if you do lots of long trips. But most private owners don't do long trips so it doesn't matter.

Administrator
Shoreham EGKA, United Kingdom

What is the basis of your argument that electronic panels are not more reliable than the mechanical gyros?

I've never had a G1000 panel go blank on me, or a failure of the ADC/AHARS, except in the sim, but that was intentional. In contrast we had quite a few problems with our gyros on the Caravans, mostly the ADI or HSI. They would last less than 100 hrs sometimes... Could be that particular set-up, King-gyros, but nonetheless, not very reliable...

ESSB, Stockholm Bromma

I have done nearly 200 hours in initially Meggitt then Avidyne with GTNs. All long trips. No failures.

And on aircraft costs, aren't you ignoring development costs?

EGTK Oxford
Sign in to add your message

Back to Top