Auto TAS Calibration with GPS

vlittle

Active Member
Joined
May 7, 2006
Messages
538
Now that you can take GPS data for the HSI page, I think it's quite possible for Dynon to have auto calibration for airspeed.

For example, one could fly (at least three) headings under screen prompt for the EFIS to determine airspeed errors. This could be done at several indicated airspeeds to build an internal airspeed correction table. This would minimize pitot-static errors (for airspeed only, not altitude).

It would also fix the 'always a headwind' problem with using GPS groundspeed and Dynon calculated true air speed. For example, if the EFIS always indicates 4% low in TAS (as mine does), no matter what direction I'm flying in still air, it would always indicate a headwind.

It would be nice if this calibration table can be uploaded to a laptop for safekeeping and analysis. Then, any changes to the pitot/static system can be analyzed. Also, the serial output stream should output corrected IAS.

I'd be happy to beta test such a feature. I know Kevin Horton monitors this list and may have some suggestions.

Vern Little
RV-9A
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
We're looking into these sort of features, but don't have anything concrete to report at the moment. We'll be in touch if/when we get to the beta test point. Thanks!
 

khorton

New Member
Joined
Nov 14, 2005
Messages
156
Location
Ottawa, Canada
I'm certainly interested in eventually having the EFIS do static system error corrections.  I would be happy to work with Dynon in an effort to come up with the best way to make this happen.  I hope to have my aircraft flying by summer, and would be interested in doing beta testing of this functionality.  

I've worked full time as an engineering test pilot since graduating from test pilot school in 1988, so I believe I could provide assistance.  I could probably entice one or two of our flight test engineers to provide advice too.

I think the first step would be to simply have a means for the user to define airspeed errors at several different speeds.  The EFIS would interpolate between those speeds.

The next step would be a way for the user to define the errors as static system error corrections, which would be applied to both airspeed and altitude.

It would be a difficult task to automate the whole process.  Any given test point could be in error, so several test points on different flights must be flown at every speed to acceptable quality data.  This greatly complicates the task of automating the whole process.  I'm not saying that this is impossible, just very difficult, and maybe not the most practical solution.
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
Vern and Kevin,
We've thought of calibrating TAS, but never IAS. Do you mean an IAS calibration too?

If so, this seems pretty non-standard to me. In an airplane, pitot pressure equals an indicated airspeed. If we change that relationship, the EFIS won't match any other instrument you put in your plane. It means that if you replace your EFIS you'll need to re-calibrate everything, or if you switch vendors of equipment in the future, nothing will match up. It removes the ability for us to calibrate a system in our shop as we build them.

I think calibrating TAS to make the winds and other calculations work out is a safe, workable idea, but I think changing IAS causes lots of problems. Calibrating static takes it to a whole new level- what's your altitude reference when you're in the air to calibrate against?

Happy to be convinced otherwise, but for now TAS was what we were thinking about.
 

vlittle

Active Member
Joined
May 7, 2006
Messages
538
I would settle for just TAS calibration and uploading the error table to allow pitot/static adjustments. It would be nice to enter correction factors for IAS, but I understand the inherent problems of mucking with primary flight data.

I know you guys aren't looking for creeping features, but having TAS be accurate is important for a bunch of reasons (navigation, fuel management, calibrated bragging rights).

And since you're listening... how about a sliding redline on the airspeed tape that uses TAS? Vne depends on TAS, not IAS. Another reason to have accurate TAS.

Vern Little
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
Ray,

Flutter is TAS dependent, not IAS dependent. This means on most planes, your real Vne is limited by TAS, not IAS. Most pilots don't know this because Vne is always shown as a mark on the IAS needle, but it can be a real issue at high altitude when you point the nose down.

Van's has a good article about this:

http://www.vansaircraft.com/pdf/hp_limts.pdf
 

lucaberta

New Member
Joined
Apr 22, 2006
Messages
30
Location
Milano, Italy
Absolutely true, especially on composite-made airplanes like my Pipistrel Virus. Infact, the Vans article specifically mentions the Pipistrel Sinus, same exact plane as mine but with a 50ft wingspan instead of 41ft of the Virus.

There have been two accidents, which luckily turned out to be non-fatal due to the use of the ballistic parachute, where pilots flying at a "normal" IAS were infact waaay fast in terms of TAS, and their planes suffered flutter-related catastrophic in-flight breakup for this reason.

The Sinus and Virus AFM specifically talks about the accidents and what to NOT do, especially when flying at high altitude, if anyone cares to read check pages 80 and 81 in the Virus AFM:

http://www.pipistrel.si/files/Virus%20ENGmanual%20Apr2006.pdf

Ivo Boscarol at Pipistrel in Slovenia also installs Dynon on request, and the display of TAS on the EFIS screen is a very good to thing to have! :)

Ciao, Luca
 

khorton

New Member
Joined
Nov 14, 2005
Messages
156
Location
Ottawa, Canada
Vern and Kevin,
We've thought of calibrating TAS, but never IAS. Do you mean an IAS calibration too?

.....
Happy to be convinced otherwise, but for now TAS was what we were thinking about.
To calculate TAS, you need to know the CAS, pressure altitude and OAT.  Errors in any of those will lead to errors in the calculated TAS.

CAS - CAS differs from the IAS seen on the ASI (or EFIS) due to instrument error and static system position error.  And, if you have an EFIS, maybe there could be errors in the algorithms used to calculate airspeed given static and pitot pressures.

Pressure Altitude - subject to instrument error and static system position error.  But, the error is usually less than 100 ft (although I have seen a report of over 150 ft altitude error due to static system position error), and an error of this magnitude has a negligible effect on the TAS.  A 100 ft error in the pressure altitude would lead to about a 0.3 kt error in the TAS at typical RV cruise speeds.

Temperature - subject to instrument error, errors in the assumed recovery factor and position error.  By position error, I mean that the position chosen may lead to errors in the temperature.  For example, many RV builders put the OAT probe in one of the NACA scoops on the forward fuselage.  This puts the body of the probe in the cockpit, so it would be heated from the back.  I have witnessed one installation where the OAT indication would vary several degrees depending on the setting of the cockpit heat control.  And I have read many reports from other builders who found errors in the OAT.  At RV cruise speeds, a 5 deg C error in the OAT would lead to about a 1.6 kt error in the calculated TAS.

The OAT indication is affected by ram temperature rise, and the probe's recovery factor should be used to remove this error.  Depending on the probe, and how it is mounted, the ram temperature rise could be up to 4 deg C at RV cruise speeds.  The amount of ram temperature rise depends on the mach number (or TAS, depending on which formula you want to use).  Thus errors in the CAS due to static system position error will lead to error in the OAT, leading to further errors in the calculated TAS.  This is a second order effect though, so the additional error is quite small.

Note that all the factors that are needed to calculate TAS are affected to some degree by static system position error.  On large aircraft, they use air data computers to remove the effect of static system position error, and this correction is applied to altitude, airspeed, mach, TAS and OAT (admittedly, some installations only apply the static system position correction to the displayed altitude).

It is probably true that most Dynon customers are only worried about the effect of all the above errors on TAS.  They aren't asking for corrections to IAS and altitude.  But, this is likely because they are not aware of the potential for errors here, nor are they aware of the consequences.  

I have communicated with one RV-4 owner who had an aircraft with flush mounted static ports.  He set his altimeter to the field elevation one day, took off and then did a high speed pass down the runway.  He noted that his altimeter read about 150 ft lower than the field elevation, and was puzzled by this.  As an experiment, he glued domed rivet heads with holes in them over his flush static ports.  He found that now his altimeter compared well with the field elevation when he flew down the runway, and his cruise IAS had increased about 10 kt.  He had been flying for quite some time with this condition, and was not aware that he was probably often flying about 200 ft higher than his assigned altitude, solely due to static system position error.  Altitude errors of this magnitude could be significant safety issue if flying IFR (ILS minima are typically 200 ft), or in VFR in airspace with IFR traffic around (the error eats up quite a bit of the assumed 500 ft between IFR and VFR cruising altitudes).

If so, this seems pretty non-standard to me. In an airplane, pitot pressure equals an indicated airspeed. If we change that relationship, the EFIS won't match any other instrument you put in your plane.
Yes, if the aircraft also has conventional round dial ASI and altimeter, having the EFIS show corrected airspeed and altitude would lead to differences between the various displayed data.  But, is really a good reason to display erroneous data, just so it agrees with other erroneous data?

It means that if you replace your EFIS you'll need to re-calibrate everything, or if you switch vendors of equipment in the future, nothing will match up. It removes the ability for us to calibrate a system in our shop as we build them.
There should be a way to backup the static system position error corrections to a connected PC, and load them onto a new EFIS if desired.  This would be no different than if the corrections were just to TAS.

It removes the ability for us to calibrate a system in our shop as we build them.
The EFIS would leave the factory will corrections set to zero, or disabled via a configuration menu.  So the existing calibration routines would still work.

Calibrating static takes it to a whole new level- what's your altitude reference when you're in the air to calibrate against?
There are various ways to do this.  For one way, see http://www.kilohotel.com/rv8/rvlinks/ssec.html

If you are interested in going down the road of correcting for static system position error, I can track down detailed info on how these corrections are normally mechanized in air data computers.  I have some contacts at a major manufacturer who should be able to give me more info on the details.
 

vlittle

Active Member
Joined
May 7, 2006
Messages
538
Kevin, thanks for the support on this. I knew you'd be on this like a Terrier.

I'm sure Dynon is asking "how much more revenue can we generate with this feature". Well, it's an arms race out there... if other EFIS manufacturers implement it first, then Dynon would be at a disadvantage. Therefore it makes sense for Dynon to be proactive on this.

I wouldn't have asked for the feature if I didn't think it was important. To have accurate TAS is important for wind calculations and Vne indication.

Your suggestion to correct for static position error is interesting. I know that TC will be forcing us to have annual pitot/static calibration soon in amateur-builts.... but this seems rather arbitrary since we may have static position errors that are not measureable during calibration. It's possible to calibrate on the ground to +/-50 feet, but have hundreds of feet of dynamic error.

NavCan is moving towards mandatory modeC in the entire Vancouver area, and is recommending traffic monitors on all VFR aircraft (Monroy, Zaon etc.). I think the sleeping giant here is that static position error may become a significant safety issue for amateur builts in crowded airspace. If Dynon can support your suggestions on static calibration this might push regulators to recommend EFIS systems that support this feature.

Vern
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
At Dynon we will always have a significant advantage over the competition- price and support ;)

I glanced at Kevin's link and didn't see where you get an accurate altitude reference while you're up in the air. GPS doesn't work- it has lots of errors, some correctable, some not. I'm still not really following how you do a good static calibration without dragging something 200' behind your plane like they do for certification. Are you just pulling the static error out by assuming all the error in the CAS comes from static position error and the pitot is always perfect?

In the end, isn't it better to actually fix your static system to actually be static than it is to try and build a table that fixes it? It still seems weird to me to fix it in the EFIS given all the other stuff in a plane that uses static. Are the Garmin G900X systems able to calibrate your static? Are they even cal'd for the airframe when installed as a G1000? I always assumed Cessna used a real static location instead of calibrating it out electronically.

In the end, I still think TAS calibration makes a lot of sense and is easy to automate, but I'm not so sure about static / IAS calibration. I have my doubts that many of our customers would go through the whole process required to make it work. Even worse, we'd have to support it ;) Maybe we can give you a table that you can manually enter corrections into, but at this point I doubt we'd be able to give you an automated process for IAS / static cal.

Just to help us out: The table would show IAS vs pressure error in PSI, right?
 

khorton

New Member
Joined
Nov 14, 2005
Messages
156
Location
Ottawa, Canada
I glanced at Kevin's link and didn't see where you get an accurate altitude reference while you're up in the air. GPS doesn't work- it has lots of errors, some correctable, some not. I'm still not really following how you do a good static calibration without dragging something 200' behind your plane like they do for certification. Are you just pulling the static error out by assuming all the error in the CAS comes from static position error and the pitot is always perfect?
Yes, this is it exactly.  Pitot errors are very, very small, unless the pitot tube is installed in a very strange place (e.g. in the prop wash, or in the wake of something else, or in the boundary layer), or significantly misaligned with the airflow.  So, you do a ground test to determine the ASI instrument error.  Then you do a flight test at a number of different airspeeds, using GPS data to determine TAS.  You take the TAS and calculate CAS.  The difference between CAS and IAS = instrument error + static system position error.  Once you know the amount of airspeed error due to static system position error, you can calculate the amount of pressure error in the static system.

This takes a whole bunch of very careful flight testing.  Each airspeed needs to be flown several times on different flights.  But you would need to do the exact same flight testing no matter whether you only wanted to correct TAS error, or correct the error at the static pressure level.

Vern - your comment on periodic static system checks is important.  I hadn't thought about this before, but it would be necessary to have zero correction to the static pressure when doing the periodic static system accuracy tests.  It would probably be good to have a configuration menu item to allow any static system error corrections to be enabled or disabled.

How many customers would go to the trouble to do the flight testing to do TAS or static system error corrections?  Probably not too many.  And I bet that some of those would cut corners and not repeat the tests on multiple flights.  So they would possibly input corrections that were not very accurate.  In this context, it might better to only make corrections to TAS.  An erroneous TAS isn't too important.  An erroneous airspeed or altitude indication is more important.
 

khorton

New Member
Joined
Nov 14, 2005
Messages
156
Location
Ottawa, Canada
I only had a bit of time last night, so I didn't answer all your questions.


In the end, isn't it better to actually fix your static system to actually be static than it is to try and build a table that fixes it? It still seems weird to me to fix it in the EFIS given all the other stuff in a plane that uses static. Are the Garmin G900X systems able to calibrate your static? Are they even cal'd for the airframe when installed as a G1000? I always assumed Cessna used a real static location instead of calibrating it out electronically.
Yes, I agree 100% that it is far better to have an accurate static source than it is to try to make corrections for the errors.  And if builders actually had accurate static sources, and accurate OAT info, then there would be no need to even consider corrections to the TAS, as it would be correct to begin with.

Van does provide an apparently reasonably accurate static source in the plans, but many builders are offended by the "ugly" pop rivet static source.  They decide to "improve" things by purchasing a flush static source.  Then, if they go to the trouble to check the airspeed accuracy, they find that they have a significant error.  Their choices now are to either rework the static system or live with the error.  In the future, maybe they would have the choice to correct for this error on TAS.  It would also be nice to have an optional correction for the error on IAS and altitude.

If someone designs their own aircraft, or if the designer has not done his homework, then there is a very good chance that they will not have an accurate static source.  It is a big job of trial and error, with several test flight for each configuration, to try to find a place for an accurate static source.  For these guys, it would be a lot easier to use the static source as-is, and correct the airspeed and altitude errors in the EFIS.

Are the Garmin G900X systems able to calibrate your static? Are they even cal'd for the airframe when installed as a G1000? I always assumed Cessna used a real static location instead of calibrating it out electronically.
The G1000, G900 and G600 use the Garmin GDC74 Air Data Computer.  I know that the GDC74B version is capable of correcting altitude for static source position error, because the Cessna Citation Mustang is designed for RVSM airspace, which requires a very accurate altitude source.  I don't know if the lower end G1000 installations use the static source correction capability or not.  The POH for the Cessna 182Q that I fly once in a while shows that despite Cessna's efforts for an accurate static source, that there are significant errors at low speed.

Maybe we can give you a table that you can manually enter corrections into, but at this point I doubt we'd be able to give you an automated process for IAS / static cal.
 
Just to help us out: The table would show IAS vs pressure error in PSI, right?
A table to make optional manual corrections would be perfect, in my opinion.  And a configuration menu to enable or disable the corrections, to facilitate periodic static system checks.

As far as what the table would have, I think the correction would be a percentage of static pressure, as a function of airspeed.  But my technical notes on this are at work, and I'll be on the road until January 4th.  I can look this detail up and get back to you then.
 

vlittle

Active Member
Joined
May 7, 2006
Messages
538
At Dynon we will always have a significant advantage over the competition- price and support ;)

I glanced at Kevin's link and didn't see where you get an accurate altitude reference while you're up in the air. GPS doesn't work- it has lots of errors, some correctable, some not. I'm still not really following how you do a good static calibration without dragging something 200' behind your plane like they do for certification. Are you just pulling the static error out by assuming all the error in the CAS comes from static position error and the pitot is always perfect?

In the end, isn't it better to actually fix your static system to actually be static than it is to try and build a table that fixes it? It still seems weird to me to fix it in the EFIS given all the other stuff in a plane that uses static. Are the Garmin G900X systems able to calibrate your static? Are they even cal'd for the airframe when installed as a G1000? I always assumed Cessna used a real static location instead of calibrating it out electronically.

In the end, I still think TAS calibration makes a lot of sense and is easy to automate, but I'm not so sure about static / IAS calibration. I have my doubts that many of our customers would go through the whole process required to make it work. Even worse, we'd have to support it ;) Maybe we can give you a table that you can manually enter corrections into, but at this point I doubt we'd be able to give you an automated process for IAS / static cal.

Just to help us out: The table would show IAS vs pressure error in PSI, right?

I agree-- Dynon's support has always been very good. This topic has been a good example.

The triple net of this discussion is that the correction of TAS is the safest and most useful function. Correcting for static position error sounds more hazardous/error prone (as Kevin discussed). Individuals can look at the TAS correction table and experiment with the static system if they wish, at their own risk.

I was thinking that the correction table would be in knots or mph.

Vern
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
I was thinking that the correction table would be in knots or mph.

Agreed if we're talking TAS, but I'm trying to wrap my head around fixing static error for both altitude and IAS. Static error should be a specific pressure error at a given speed, and this pressure error leads to an IAS error (worse at low speeds due to the non-linear nature of pressure vs speed) and an altitude error (also non-linear).

The real trick is the non-linearity. If your static system is .1 PSI low at 100 kts IAS, the error in feet will increase the higher you go, and the error in speed will decrease the faster you go, so you need to correct it in the pressure domain, not in the feet or knots domain. If all you correct is IAS or TAS, you can do it right in the speed domain since it's all related.

I'm starting to think Dynon should sell a pitot with a static source on it that is 25' long and sticks way in front of the plane!
 

JetJockey

New Member
Joined
Nov 2, 2006
Messages
38
Location
Granbury,
Flutter is TAS dependent, not IAS dependent. This means on most planes, your real Vne is limited by TAS, not IAS. Most pilots don't know this because Vne is always shown as a mark on the IAS needle, but it can be a real issue at high altitude when you point the nose down.

You are correct except that Vne can (and usually is) limited by dynamic pressure, i.e. IAS. My point was that the Vne is an IAS limit, whereas Vmo/Mmo are TAS limits. I seriously doubt that there are that many customers that would really benefit from you doing the extra programming to provide a sliding Vmo/Mmo bug - but that is just my opinion.

Regards,
Ray
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
According to the article by Van, the RV series is limited by flutter, not dynamic pressure, and thus the real thing you need to keep track of is TAS. Agreed that Vne is a number represented on the IAS indicator, but the true number you want can often be below that, so it does seem like it would help a lot of people out.
 

JetJockey

New Member
Joined
Nov 2, 2006
Messages
38
Location
Granbury,
I guess all I am trying to get across is you need to call the limit you are referring to the correct thing - Vmo. I guess I don't really see a scenario where a RV could attain an altitude where the IAS/TAS split would be so significant that a reasonable Vne would not provide a sufficient margin of safety, but if you want to expend the effort to incorporate this feature, more power to you!!

Regards,
Ray
 

dynonsupport

Dynon Technical Support
Staff member
Joined
Mar 23, 2005
Messages
13,226
Ray, if you read that article by Van, you'll see that an RV-4 at Vne at 11K feet is 40 knots over the Vmo. While it can't cruise at Vne, it doesn't take much of a nose down attitude to hit Vne, and due to the TAS at that altitude, it's way over Vmo. I'm not sure most pilots want to assume that the Vne recommended by Vans has plenty of overhead in it no matter what their altitude.

Agreed that few piston airplanes can cruise above Vmo, but gravity has a lot of horsepower ;)
 

jim

New Member
Joined
Nov 20, 2006
Messages
21
It would seem the best way to avoid calculated TAS errors would be to ensure you are using Calibrated airspeed instead of indicated airspeed (with all its errors) in TAS calculations. I would like to see Dynon offer a correction of IAS -> CAS at Stall/MCAS/approach speeds, and one at cruise speeds. I think winds are of concern during 2 phases of flight: cruise & approach. It makes sense to have calibrated airspeeds for at least these two speed regimes to enable accurate calculation of TAS & winds.
 
Top