Jump to content
Sign in to follow this  
Guest Jasonfly

RNP modelling

Recommended Posts

Guest D17S

Your question was . . . "Is PMDG's LNAV/VNAV good enough to excute such approach assuming we just want it to fly through all the waypoints and not worry about a radius for a moment?"Sorry I wasn't clear. The answer is "Yes."

Share this post


Link to post

I set up the KDCA RNAV RNP RWY19 approach in the SIDSTAR file, and made three attempts at it from KBWI using the BELTS transition. See attached approach plate.The PMDG FMC failed to execute the approach. The first time I let VNAV/LNAV try it, then I used LNAV for trackig and controlled speed and descent via the MCP. For the 3d attempt, I inserted a fix about midway between KUBEY and HIXIT to approximate the arc between these two fixes but it also failed.The reason for failure was the FMC's inabiiity to capture a desired track in time to deal with the next change in track following each fix. The simulation flew a series of S-turns that increased in error amplitude as the distance between fixes decreased.I think the reason RF-legs (radius to fix) are used is to ensure that the aircraft is on a specific track at each fix. For example, there is a 115-deg change in heading between KUBEY and HIXIT. Leaving KUBEY, the desired track turns left from a 260-deg course to be on a 145-deg course when it reaches HIXIT. Even with a mid-arc fix, the PMDG FMC was unable to get close enough to a 145-deg course on reaching HIXIT to intercept the track and be on a 145-deg course by the time it reaches the next fix JTSON (FAF). The error amplified as the distance between fixes decreased between the FAF and RWY. The aircraft never got close to the the runway.The obvious next step is to insert more fixes on the desired track to see if the intercept errors can be reduced. I'm thinking about trying the technique used in the "How to Design" tutorial for DME arcs. I guess this problem has hooked me and I'm gonna see it through to a definitive conclusion. There are a couple of RNP approaches in KPSP that I'd like to try as well, but first we gotta figure out the method.FYI: A RF leg is defined by four variables: Two fixes, the course leaving the first fix and the track radius. The course leaving the first fix is the tangent of the arc at that fix, and the radius is the distance to the center of the arc from the fixes. It is easy to calculate the track give the four variables, but the chart doesn't provide the radius. It provides two fixes, the course leaving the first fix, the course arriving at the second fix, and the approximate arc length. I need to recall my high school geometry to solve the problem.Give me a couple of days. If anyone else wants to try it I can post the fix coordinates.


Dan Downs KCRP

Share this post


Link to post

Dan,Good effort on your part.Are you missing the formula for the radius?I can give it to you but you have to realize this formula assumes it is possible to draw an arch of constant radius between those points. In the most general case of two points and two courses there is no guarantee that such radius will exist. But assuming it does this is the formula:R = D/(2*sin(a))D is the distance between those two points, 'a' is the angle between the course in the first point and the course drawn between the two points. So for example if the course at the first point A is 30 and the course line from A to B is 340 the angle 'a' will be 30 - 340 = -310 = 50 deg.It is easy to verify if conditions for the constant radius are met. This number that I calculated in the example above (50) should be the same if you did those calculation in the point B. In other words there exists certain symmetry between the course in A and B and the course connecting A with B. Michael J.http://www.precisionmanuals.com/images/forum/pmdg_744F.jpghttp://sales.hifisim.com/pub-download/asv6-banner-beta.jpg

Share this post


Link to post

Michael: Thanks for the assist. I found very similiar relationships in my old CRC handbook, and I re-discovered Aviation Formulary at http://williams.best.vwh.net/avform.htm (a great resource).I'm still playing with it. I have built the arc between KUBEY and HIXIT, which took several hours to cook up. However, after the first one the rest were fast food. I'm still tweeking, but I think the results are going to be better than I expected. I've got it following the path all the way to the DH in LNAV. Still need to figure out what happens after that, it is almost as tho the RNW19 fix in the SIDSTAR is off to the east by several hundred feet. I have no idea where that fix is generated.I am also unhappy with the VNAV performance, but I haven't given up on that yet.I'll attach the text that can be used in the kdca.txt file but I will assume the you or anyone else that uses it knows where to put the fixes and how to replace the old RNV19 with the new. Be sure to backup original.More to come...


Dan Downs KCRP

Share this post


Link to post

This is my final report on the attempt to build a flyable RNP approach. I have attached the text that can be used to add this procedure to your SID STAR KDCA.txt file. First, make a backup copy then copy and paste the fixes into the fix section and replace the outdated RNV19 approach with the RNV19 approach from my file. The approach includes all transitions.Can the PMDG FMC fly a RNP approach? The answer is yes, but.The problems are:1) VNAV conditional fix bug interferes when maintaining VNAV Path is important. This is a known bug that is listed on the maintenance board but PMDG isn't fixing it. This is the one where the VSI sticks to a purple bug and quits following VNAV Path.2) VNAV speed control will not follow the deceleration profile that the FMC selects after pilot selects approach flaps. For example, I designed the approach to slow to 210 at IAP but manual MCP Speed control seems to be required to slow it down futher for FAF and final descent -- but manual speed turns off VNAV path, so here's another reason for not having VNAV.Besides not having any VNAV for the approach, the LNAV is not capable of RNP.11 (it is probably alright for RNP2.0). Most of the approach is flown by LNAV pretty faithfully (I had to add many many custom fixes to get the curves just right), but it consistently fails to keep the aircraft out of P-56 (the Capitol, Mall etc) and that is why this approach was written to .11nm error instead of the normal .3nm approach error. This rule break always occurs, now matter how I reconfigure the overfly rules and adjust custom fixes. It happens on what should be a straight path between JUBOL and WIRSO, but the LNAV is not following the track tightly enough.This is a real tricky approach to fly because you must attend to the VNAV errors, keep track of descent rates and fix crossing altitudes, get everything configured for final descent right before JTSON, and then the challange becomes handflying the curving final mile over the bridges and lining up on runway centerline (which happens to be blocked by the centerpost during the final turn). The RNAV approach will not fly the airplane to the runway like an ILS, nor does it provide autoland. You gotta handfly this one for the last mile as you descend below 475 MSL. All in all a lot of fun. I've made it to the runway about half the time.Bottom Line: VNAV not working good enough. LNAV probably good enough for RNP2.0 and maybe RNP0.30 but never RNP0.11.This is fun!


Dan Downs KCRP

Share this post


Link to post
Guest D17S

Dan, I agree. So I picked up your updated KDCA, RNP approach for runway 19 from the flight plans area. It worked perfectly with the 744. There was never a problem staying clear of the P-56 restricted area, however I kept slipping wide on the final turn to the runway. Once I lightened my GW to < 500Klbs, the airplane finally started making even the final turn in good order. My

Share this post


Link to post

Michael-You and I have the same reading habits. :-)You still flipping to the back of Flying before reading the front? hahahahahha....For those interested- we do not intend to model RNP mathematics at this time....It's what we call "beyond the scope" of what we are looking to accomplish....The PMDG 747-400 flies LNAV/VNAV more precisely than anything out there... Not much benefit in tightening it yet....


Robert S. Randazzo coolcap.gif

PLEASE NOTE THAT PMDG HAS DEPARTED AVSIM

You can find us at:  http://forum.pmdg.com

Share this post


Link to post

Sam: I don't mind at all. I appreciate your interest.The RNP procedures are going to appear in more and more places, especially SIDS/STARS. For example, KDFW is up on running on RNAV SIDS, which while not defined as a RNP proceedure, they do provide RNP 2.0 instructions for so equiped aircraft. An article in Aviation Week and Space Technology highlighted the many benefits including significant improvement in keeping everyone on the same tracks, fewer radio transmissions and increase traffic rates compared to radar vectors. I think it is the least cost method for the FAA to increase capacity without having to add controllers or runways.Thanks again for your interest.PMDG's attitude is understandable but regrettable. I sure wish they would get rid of the soft conditional bug in VNAV and add RF leg capability to their FMC simulation.


Dan Downs KCRP

Share this post


Link to post

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  
  • Tom Allensworth,
    Founder of AVSIM Online


  • Flight Simulation's Premier Resource!

    AVSIM is a free service to the flight simulation community. AVSIM is staffed completely by volunteers and all funds donated to AVSIM go directly back to supporting the community. Your donation here helps to pay our bandwidth costs, emergency funding, and other general costs that crop up from time to time. Thank you for your support!

    Click here for more information and to see all donations year to date.
×
×
  • Create New...