Speed Intevention Question (HPSOV?)

[Guest Ian_Riddell]

"In PIC VNAV does not infact use idle, but cheats a little, which may well mean that speed intervene will work (as a side note speed intervene DOES work on the 744, and our newest Pegasus FMC 767, although not on our other 2 pegasus 767's...)."Just been asked a question on another forum regarding speed intervention and wanted to confirm what the indications/aircraft behaviour would be on the 744 and newest Pegasus FMC in the following circumstance (see diagram):http://members.ozemail.com.au/~b744er/767/...ntervention.GIFA) A large speed intervene at point "A" in idle descent causing the aircraft to deviate from the planned vertical flight path (shown in purple): Would the deviation indication on the EHSI show the aircraft above the path?:( A speed intervene to 160 kts (20 kts greater than the planned 140kts at the approach to the runway) at point "B": Would the aircraft add thrust to maintain the original VNAV path & would the deviation indication on the EHSI show "on path"?The gentleman who asked me tells me that he has to Speed Intervene in all his descents (as his aircraft will not slow down unless he does this). He appears to be training on the 777 which, I assume, is similar to a 747-400 or Pegasus 767.Thanks.Cheers.Ian.

[Guest]

I will write what I know and hopefully it helps in some areas for clarification of the programming for Vpath logic as I know it.On the FMS CDU, the pilot has entered the following parameters to fly a Vpath descent:1) Bottom of descent (BOD) altitude at waypoint2) An angle of descent if the pilot wants to fly a particular flight path angle that is different from the one that was programmed into the FMS prior to the days flight. ( PIC I beleive has an angle defined based upon the descent target speed and BOD )All of the other required VNAV parameters are provided as inputs to the FMS navigation computer or are called up automatically from the FMS navigation database.In the the navigation computer, all of these parameters are combined to generate an FMS computed descent angle. The descent angle is converted to vertical speed, using ground speed as an input. The commanded vertical speed is then routed to the flight director.In the flight director, the FMS commanded vertical speed is summed with actual aircraft vertical speed. The difference between the two vertical speeds is the command the aircraft follows.Based upon the above information, If the descent target speed is set and the angle of descent is fixed and the vertical speed is computed, the only thing the aircraft will follow is the change in vertical rate and not speed changes. If you have a late descent, the deviation indicator will show your position above the path on the MAP display. Since the path is below your position, the FMC engages in VNAV SPD instead of VNAV PATH. The airplane will descend at the descent target speed and may remain above the path unless the pilot intervenes with a steeper descent rate and /or drag(speed brakes). If speeds are changed with out recomputing the VPATH requirements , which also means resetting the descent target speed, you will not always be able to follow the path as computed.Multiple segment descents may be defined but it is only the first segment that has an idle path trajectory. Subsequent descent paths may or may not be idle depending upon steepness of the path.The FMC bulds the vertical path starting at the lowest wpt constraint and then projecting up to cruise altitude. I hope this helps from my little bit of knowledge aquired over the years.Regards,Cliff Scott

[Guest Ian_Riddell]

>In the the navigation computer, all of these parameters are >combined to generate an FMS computed descent angle. The >descent angle is converted to vertical speed, using ground >speed as an input. The commanded vertical speed is then >routed to the flight director.This seems a little different to my understanding, Cliff. I thought that the descent was plotted as a 2 dimensional path in space (based on aircraft ground position and altitude)... a bit like a glideslope... and the autopilot simply tries to capture/follow the "glidepath". E.g. at Xnm from the lowest waypoint constraint, I should be at Y feet. An ideal airspeed is computed to follow this vertical path, but, as we know, in VNAV Descent, the aircraft can increase it's airspeed up to 15kts (not exceeding Mmo/Vmo) to recapture the path (if the aircraft is high). Any more and it has to leave the path. >In the flight director, the FMS commanded vertical speed is >summed with actual aircraft vertical speed. The difference >between the two vertical speeds is the command the aircraft >follows.I was thinking that, in idle descent, it was more a case of the FMC commanding an airspeed (through elevators) to capture a path in space, rather than the FMC commanding a vertical speed?Some of this stuff gets a little beyond me at times ;-) Thankfully it falls under the category of "nice to know"... rather than a "need to know" for the type of work I do.Thanks.Cheers.Ian.

[Guest]

Hi Ian,My position is as a system specialist for avionics and electrical systems and I have the fortunate opportunity to work with a guy (top notch) who was part of the origanal design team for the Boeing FMS system and currently the Gulfstream FMS system.My specialty is FMS, Autopilot, and Flight Director systems to name a few. I work directly with Field Service on systems related to the Collins, and Honeywell names. The internal function for determining VPATH angle has not changed over the years from the different systems. Yes, airspeed is brought into the equation as a means to measure deviation because +/- speed and aircraft dynamics (lift properties) will cause the aircraft to depart the path. It is also true that increased speed on the descent will drive you to the path if you are above it, but the vertical rate is what drives the airplane on path. You have to remember speed is a fixed value for the path derived from the descent target speed in the FMS. This speed does not vary during the descent phase in VPATH. The system has the capability to control the vertical rate within certain parameters but when the aircraft gets too far from the computed path descent rate, it will go out of VPATH as indicated earlier. In the Boeing and Gulfstream, but not the PIC system, wind correction is also part of the computation as long as you are not descending from the Jet Stream. When I fly my Cherokee 180, on the G/S or a descent from altitude to a fixed waypoint at a lower altitude and want to keep a fixed angle of descent, I contol my vertical speed (speed on pitch) and hold my airspeed at a fixed value (RPM setting)Regards,Cliff Scott

[Guest]

Hi Ian,I called my friend from Honeywell and got further clarification from him on the FMS operation and also flew the sim again to see what the mode transitions were in relation to system performance.There is a mode not indicated by the sim which is called VFLCH mode but is the same as when you are in VPATH and IDLE A/T. This is Speed on Pitch to correct for angle of descent errors (vertical speed rate). When the Flight Director mode changes to N1 A/T it is Speed on Throttle with about 8 deg of PLA movement to allow speed intervention to correct for going off the VPATH, but this mode is controlled by the performance computer and may not hold VPATH depending upon the descent angle whereby the A/T would go into IDLE A/T (Speed on Pitch) to compensate for a steeper desc angle. I flew thr PIC sim and duplicated the modes.I was at fl350 desc to 10000 and over gross to get an error in performance. On initial descent the Flight Director went into VPATH and N1 A/T and controlled the aircraft with throttle. As the aircraft started to go off the path the Flight Director stayed in VPATH and the A/T went into IDLE A/T mode to increase the desc rate( vertical speed), in which the desc rate did increase to capture the VPATH.(note the aircraft did pitch up and the command bars on the flight director showed commanding a nose up position). In the process of increasing the desc rate you will also have a tendency to increase speed and more than likely bust your restriction. Normally you would get a message telling you to increase drag to bring the aircraft back on speed and path. PIC does not give this command but at that time you would need to deploy the boards. During IDLE A/T you do not have Speed on Throttle correction. I do not normally have a problem holding speed or VPATH as long as the aircraft performance falls inside the FMS limits as prescribed in the software for PIC. This would mean not to overgross the aircraft where the computed figures would be invalid.It is concidered good practice if you have control of your descent to engage VNAV desc about 50nm prior to the computed TOD to shallow out the desc rate to help stay on speed and maintain a desc rate of 1200 feet/min. If you cannot set this and ATC is controlling your descent then use FLCH which is Speed on Pitch and easier to control your descent especially below 10000. I find I am using FLCH mode quite a bit during the busy times and do not bust my speeds.Regards,Cliff Scott

[Guest MB327]

Yea, and E=MC Squared..LOL, just kidding there Cliff. I had to read that 5 times and I am still brick shy of a full load of understanding it. Very interesting to read analysis of the VNAV system. These kind of contributions, along with Ian and HPSOV, are most welcome on the forum.

[Guest Ian_Riddell]

Thanks for your help/clarification, Cliff.It may take me a while to analyse the full contents of your message (Mike is not alone here :-))... I may have to read my current FMC "bible", the Big Boeing Users Guide by Bill Bulfer/###### Gifford a few more times also, to see if I have misinterpreted something.On perhaps a simpler subject....I'm not sure why you would need to make an early descent. Does this shallow path, with the possibility of manual throttle intervention/control, enable you to better maintain your airspeed down to a lower altitude (giving the FMC less chance/time to make a mess of the final stages of descent!). I remember being told that descending too early (i.e. beyond 50nm from TOD) may confuse the FMC, so I wasn't sure if any early descent was the best way to go.Cheers.Ian.

[Guest]

I'm not sure why you would need to make an early descent. Does this shallow path, with >the possibility of manual throttle intervention/control, >enable you to better maintain your airspeed down to a lower >altitude (giving the FMC less chance/time to make a mess of >the final stages of descent!). I remember being told that >descending too early (i.e. beyond 50nm from TOD) may confuse >the FMC, so I wasn't sure if any early descent was the best >way to go. The FMC doesn't get confused but calculates for a 1200 foot per minute descent in the FLCH mode and VNAV SPD or N1 A/T so you can have better control of the speed prior to intercept of the VPATH. The aircraft is under the VPATH and flys into it for capture of the VPATH angle. This procedure is fine if you are not controlled by ATC or can get clearance to start the desc when you want to. The fellow flying the 777 is flying at a higher gross weight than the 767 and I feel his problem is having the computer establishing values based upon the gross weight of the 777 rather than the 767, if the person you are talking about is using PIC and not the Class-D sim for the 777? The gross weight of the 767 for the PIC sim is 408.4 and if he is above gross the aircraft will never fly the VPATH due to computations based upon the 767 gross weight of 408.4. I pretty much proved this out today flying heavier than allowed gross for the PIC FMS.I also mentioned earlier that the first leg of the VNAV descent is set for IDLE A/T and the ability to go into Throttle Hold (throttle servo diconnect) but any restrictions after that may have a descent angle to great to be controlled by N1 or Spd and the boards would have to be deployed to control speed and path angle.If ATC is controlling you after the initial descent, you are better off using FLCH where you have better control of the aircraft with Speed on Pitch rather than throttle. Regards,Cliff

[Guest]

As thoughts come into my head I figure I ought to set them down in the forum for review and discussion. By doing this I also learn things that I also have difficulty understanding or have been misinformed. I also have read the manual by Bill Bulfer and and ###### Gifford. Along with that and the information from Honeywell and my experience using and working on these FMS systems I have tried to clarify my understanding of system integration and performance for everyones benefit including me. I enjoy forums like this where everyone learns.I would like to summarize some of what we have been talking about in the latter posts.On the initial leg of VNAV interecept the FMC determines VPATH descent angle based upon the input of ground speed and the BOD. With ground speed a value for vertical speed is computed for IDLE A/T or Speed on Pitch. The system starts out with VNAV descent using VNAV SPD in N1. to acquire the VPATH ( Speed on Throttle) Once the path is acquired the FMC will go into IDLE A/T or Speed on Pitch. As an example lets say the descent rate estabished for the BOD was 1300 foot per minute. The FMC will maintain this rate as the path angle. Speed on throttle will allow us to correct for any deviation from the path as long as the desent angle is not changed to the point that we cannot slow the aircraft enough to increase our descent rate to catch the path in case of flying above it. If that is the case then the boards need to be deployed to bleed off speed which will give an increased descent angle or increased vertical speed until on the path again at 1300 foot/ minute.I hope I haven't beat this subject to death but could talk on it for days.Regards,Cliff

[Guest tascaso]

Thank you Cliff and Ian...for all the brain food. I will have to hop into the old PIC767 and do a few hops and try this out. It never ends the learning!Thank youTony

[Guest HPSOV]

Well I've read all the replies so far... and to be honest most of them went over my head!From a simple pilots point of view (a simple point of view or a simple pilot?... I dunno).In example A, in a pegasus aircraft a speed intervention would cause the aircraft to change from VNAV PTH to VNAV SPD, the aircraft will pitch to obtain the speed entered in the window, and the aircraft would rise above the VNAV path, as indicated by the VTK and PROG2/2.In example B, the aircraft will be in the approach phase (approach phase occurs after the first waypoint of an FMS loaded approach, or within 25nm if the runway is the active waypoint). Once in approach phase the pitch mode remains in VNAV PTH after speed intervention, and the vertical path is kept regardless of IAS MACH selector changes. Thats the quote from the manual... having never tried this in a pegasus aircraft I cant expand anymore. It sounds as though it will operate just like a non-pegasus aircraft, by following the path.If the gentleman in question is having to speed intervene in all descents I would suggest he is not loading VNAV correctly, as there is no reason why in a properly planned descent there would be any need for speed intervention (apart from ATC or un-forecast winds etc).

[Guest HPSOV]

The manual states that you should not use descend now function when more than 50nm from TOD as the FMC fuel predictions will be inaccurate, which could cause a fuel consumption error (whatever that is).

[Guest Ian_Riddell]

"The manual states that you should not use descend now function when more than 50nm from TOD as..."Thanks HP.That was the advisory I was thinking of, but didn't know where I'd seen it."Once in approach phase the pitch mode remains in VNAV PTH after speed intervention, and the vertical path is kept regardless of IAS MACH selector changes."The plot thickens... (doesn't it always!)"If the gentleman in question is having to speed intervene in all descents I would suggest he is not loading VNAV correctly,..."That was my first impression too, but, of course, it could be that his company hasn't bothered working out the correct FF/Drag Factors for his aircraft (or Maintenance forgot to enter.. or stuffed up the entries during programming!).Anyway.... back to work...http://members.ozemail.com.au/~b744er/PVDRepair.jpgCheers.Ian.P.S. Anyone seen my screwdriver?

[Guest Ian_Riddell]

Sounds like a lot of number crunching whichever way you look at it, Cliff :-)In particular, it's the "when" part that has me scratching my head.... e.g. When does the aircraft know it's at TOD? When does the aircraft know when to start slowing before 10,000' to avoid a bust? Knowing Groundspeed is one thing, but you need a lat/long for the BOD as well as an altitude, so you can compute backwards to know when to start descending.I was thinking that it would easier to define a VNAV vertical path in space if it were more like a glideslope, rather than dealing with vertical speeds. Still, as you say, these speeds are are based on ground speeds, rather than airspeeds, so it pretty much sorts itself out in the wash. I'm beginning to understand what you are saying, Cliff ;-)Thanks.Cheers.Ian.P.S. Off topic folks, but if you think the FMC is complex, try juggling the fuel on this aircraft... 9 fuel tanks!http://members.ozemail.com.au/~b744er/744e...elSynoptic2.jpg

[Guest Lenny Zaman]

GRRRR at Ian, you hurt this girl!!!!!look at her totally banged up like that :-pLOOOOOLoh and, yes i got yer screwdriver here in Belgium.;. hues it got on a long flight to EBBR ;) ;)Nice piccies and thanks for the info