The mysteries of the height-velocity curve

Call it what you want; height-velocity curve, dead-man’s curve or even limiting height-speed envelope for those who like sophisticated phrases. The “dead-man’s curve” is probably a carryover from our fixed-wing brethren while the industry generally accepts the simple reference of “H/V curve”. The inside of the curve is the area from which it will be difficult or nearly impossible to make a safe landing following an engine failure if you are in the same conditions depicted with respect to airspeed and altitude.

The H/V (height-velocity) diagram is a staple in the helicopter arena but sadly is often misunderstood by student and instructors alike. So, let’s take a look at what it is and how it is developed.

What is it?

The Height-Velocity diagram (curve) is a chart showing various heights above the ground with a combination of a velocity (indicated airspeed) where successful autorotation and landing is or is not possible. This magical combination of numbers yields two major regions on the chart; the area above the knee and the area below the knee. These areas are what actually plot much of the “curve”.  During initial helicopter certification, test pilots evaluate several characteristics of the helicopter that help determine the H/V curve. These factors include the helicopters initial response to a power loss, steady-state descent performance and power-off landing characteristics and capabilities.

Unknown to many, the development of the H/V curve and its associated number combinations is based on “pilot minimum skill level”. So, in a perfect world this means if the engine fails while I’m going this fast (KIAS) and at this height a pilot at a “minimum skill level” should be able to make a successful autorotation and hopefully some resemblance of a landing.

How do we define the pilot “minimum skill level?” That is a question I and many others can’t answer. Many will agree the current practical test standards are somewhat lacking and aren’t necessarily cultivating the “minimum skill level” necessary.


No cookies for me please

To delve into this quandary let us recap the typical sequence of an autorotation training exercise. The instructor has the student line the helicopter up with the runway so that the power-recovery phase of the autorotation will occur as closely to the runway numbers as possible. Sound familiar? You know what I’m talking about, the “3, 2, 1, roll-off power” etc.

It’s the same thing with a 180-degree autorotation where the student is taught where to “fail” the engine based on tailwind strength and land at the “spot” within the practical test standard. Is there really anything practical about it?  Just what, if an engine failure occurs in the real world and the only spot you have is 600 feet directly below you? Could you get there? Safely? What about engine failures at night time with the same situation, the only place to go is directly below you or just out in front of you. If you remember anything from this article remember this, autorotations are like fingerprints in that no two are exactly the same.

How is it developed?

When test pilots are developing the H/V Curve for a specific helicopter the testing is done in very defined environments and conditions. The test is conducted at max gross weight for sea level and max OGE weight with the lesser of aircraft maximum altitude capacity or 7000 density altitude with 2 knots or less of wind, and with the landing portion being conducted to a smooth hard paved surface. Maybe this “worst case scenario” is what helps define the elusive “minimum pilot skill level” in the hopes that the average pilot will not often be found in these conditions.

The FAA references for H/V as used during certification.

The FAA references for H/V as used during certification.

Another tidbit of importance to mention is that when H/V curve testing is being done it is accomplished by a “throttle chop” and not an actual flameout. As best described by my friend Pete Gillies of Western Helicopters, “There is a world of difference between chopping the throttle to idle versus flaming out the engine.” In one of my favorite helicopters the engine actually develops 15-17 horsepower at idle. This minimal power can make the flare and pitch pull at the bottom so much easier than if the engine has gone away completely. The piston-powered helicopters on the other hand can deliver a more realistic prediction of what pilots may experience when they have a real engine failure. In turbine helicopters only the Lama, Alouette and Gazelle will provide similar results as the piston-powered because of their centrifugal clutches.

The development of the H/V curve also includes an “intervention time” that attempts to mimic the surprise factor encountered in real engine failures. This intervention is measured in terms of seconds (or lack thereof) depending on where you are located on the H/V curve. 

The civil certification calls for a 1-second intervention at an area above the knee. So, in essence, during testing and certification the throttle is rolled to idle and a “thousand-one” count is initiated. The standards require the pilot’s hand to be on the cyclic and feet on pedals (sounds reasonable) but the pilot’s hand is not required to be on the collective. This standard applies to your basic helicopter certified under FAR Part 27 standards. Although it is assumed you won’t have your hand on the collective above the knee the standard requires that the power be set for level flight at any specified airspeed. Certification standards for areas below the knee have NO intervention factor as it is assumed you will be at full or takeoff power with all feet and hands on controls.

Why it all matters

So how do we tie all of this knowledge together? Some agree that the theory of entering and being in autorotation isn’t necessarily lacking as most can accomplish that phase safely, it is the more advanced autorotation fundamentals that are being left out of most syllabi, such as teaching how to get to a spot by varying airspeed (and rotor RPM) while in autorotation. It is what transpires between the time of the engine failure and the point at which time the pilot initiates the flare that can make all of the difference. Getting the collective down has been hammered into students heads from the beginning, and it is obviously required and critical; however, getting the cyclic back immediately is absolutely critical!

As a designated pilot examiner it is common to see applicants struggle with hitting their spot. Why? My subjective opinion is that they are taught to enter the autorotation at a specified location in order to hit that spot. They enter the autorotation, realize they are short or long and then the specified airspeed and rotor RPM they were taught is either getting critically low or astonishingly too high. Who is the blame? The student? The instructor? Or the system? After all, it is the practical test standards (PTS) that dictates that an applicant will “Establish proper aircraft trim and autorotation airspeed, ±5 knots.”

I am absolutely amazed to see students, pilots, and instructors alike think something bad is going to happen if their airspeed drops below the “recommend autorotation airspeed” during the initial entry or at some point below the high hover point and or above the “knee”. Where does this mindset come from? Obviously it comes from the PTS that dictate that an applicant will maintain best autorotation speed within the given tolerance. This is rather unfortunate considering airspeed indicators are typically not reliable in autorotation. In an engine failure or a practice autorotation approach from cruising level altitude (at or above the high hover point on the H/V curve) all I care about is getting upward airflow through the rotor disc to serve as the driving force and making sure my rotor RPM is where it needs to be. Initially I couldn’t care less about airspeed. The airspeed becomes critical prior to the initiation of the flare for one solid cardinal principle, which is that there is an airspeed below which the flare will be spectacularly non-effective. I’m not suggesting that one become a weekend warrior test pilot.  Don’t experiment to find the lowest airspeed you need at the flare initiation. Just know you will need a practical amount of airspeed and this is something that can be learned with the proper instructor. Also keep in mind the design factors of a helicopter; unlike a fixed-wing aircraft helicopters are designed to land vertically with little forward horizontal momentum.



  1. Great read, Matt! Pete was very excited to show me a zero-speed auto once, and it really opened my eyes as to the amount of ‘tools’ that we can use in an auto, yet don’t do or aren’t taught because it’s not part of the checkride. I know my instructor would have liked to do more of these, but they were minding my time and money, which is dictated by the PTS. Full-down autos (with qualified instructors) could also be added to your list of things qualifying minimum pilot skill, since a real auto isn’t going to contain a powered recovery. I hope this article saves a life someday! Cyclic back!

  2. G’day,

    A diagram of the H/V would have been helpful.

  3. Nice article. Something you touched on and which I have argued for years is the autorotation maneuvers required by the PTS, as opposed to, a simulated engine failure (throttle chop) during training or testing. The ‘Straight-in’ and ‘180 Auto’ required by the PTS ARE NOT EMERGENCY PROCEDURES. They should not be trained that way, nor should they be practiced that way. Those maneuvers are there to teach the students the fundamentals of manipulating the helicopter, in flight, in a power-off situation. So many instructors make the practice so stressful for themselves and their students, that no one gets anything from the training except for an adrenaline surge. Those maneuvers should be made as slowly as can safely be done, as controlled as possible and as smoothly and precisely as the skill level of the pilot at the controls will allow. That is how a student learns to judge wind direction/velocity and the characteristics of their helicopter under those conditions, that will end in the successful engine failure scenario in the future. That is completely different than a simulated engine failure. There is a time and place for throttle chops, but, don’t confuse them with autorotation training.

  4. Good article. I would put most of the blame on the PTS as the author indicates. Enhanced autorotations should be taught but flying helicopters is expensive and CFIs want their students to pass without breaking the bank, so there is that balance they must consider throughout training. One thing I would add is too many CFI’s teach autos to initial students starting at 500′ AGL. The maneuver is over in less than 30 seconds so there is no time for the student to learn. Back in the day I used to take my initial students up to 3 or 4 thousand feet AGL and let them ‘fly’ the autorotation. They learn A LOT more about how the aircraft reacts to various control inputs because they have have several minutes to absorb the experience which allows them to retain the knowledge gained b/c is is a much calmer and relaxed maneuver.

  5. Thanks for providing some background that students don’t get when they are training for autorotations. Instructors discussing “what ifs” with students helps them understand the capabilities of the helicopter during an autorotation when they are usually focused on passing the checkride to the Practical Test Standards.

  6. Good article in concept however I don’t believe that all HV curves are developed at 7000′. Robinson did theirs at both sea level and 7000′ because they have Big Bear available. Remember that auto practice is just that. Learning how a helicopter acts in autorotaion to effect glide angles, rotor control, effects of winds, different attitudes, etc.

    Engine failure practice needs to be done as forced landings with no warning at various altitudes, airspeeds. location of landing areas. If your instructor does not do this, go somewhere else. We do autos from zero to VNE.

  7. Hi Matt. The H/V curve illustration that was added to your article leaves me somewhat perplexed. In the first paragraph of the “How is it developed?” section of your article, you say that H/V testing is conducted at 7,000 feet density altitude and at maximum gross weight. In the illustration I see two H/V curves, one for sea level at a weight of 2500 lb, and one for 8,500 feet at a weight of 2250 lb. Each of the helicopters with which I’m familiar has only a single maximum gross weight that is listed in the limitations section of the RFM. Does the helicopter represented by the illustration have two maximum gross weights? Also, if testing is conducted at 7,000 feet as you state, why is a curve shown for 8,500 feet?

  8. Great article! After I completed my Commercial Rotorcraft add-on to my fixed-wing ratings I asked my examiner to show me something that I might not have learned in my training… So…. He started his auto over top of our intended touchdown spot… Backed-up….. In autorotation.. Put the touchdown spot in the correct spot ahead if us.. And re-entered the forward autorotation to a normal flare.. Very impressive. Another ‘tool’ to keep in my box… Should the situation present itself someday… 25,000 hour of fixed-wing time – and helicopters never cease to amaze me.

  9. Even the PTS for CFI’s barely touches the possibilities available in autorotation, though being required to perform them to touchdown is an improvement on the private and commercial checkrides. The rotor doesn’t care whether you’re going forwards, backwards, sideways, or just descending straight down with zero airspeed, all are possibilities. Provided you manage your airspeed and rotor RPM such that you can accomplish an effective flare and touchdown on your intended spot, it doesn’t really matter what you did to get to that point. I was lucky enough to have an old Army instructor pilot show me the ropes back when I was flying Hueys in flight school, and I have to tell you that the first “backing up” auto was a real eye-opener, as were his tight spirals down to the flare.

    It’s also worth noting that autorotations (other than hovering) are actually more difficult at lighter weights, which is totally counter-intuitive. Anyone who has flown R-22s knows how different the auto and flare feel the first time you practice one solo, since the weight of the instructor is actually a pretty significant portion of the gross weight on that aircraft. If you check the maintenance manuals for most helicopters, you’ll find that the minimum collective blade pitch settings are adjusted predicated on the aircraft at minimum weight, since that’s the most demanding flight condition.

    It’s also worth noting that most, if not all, multi-engined helicopters are much more challenging to autorotate to a successful touchdown with both engines inoperative. Single engine helicopters typically have far more available blade inertia to use at the end of the flare relative to their max gross weight, and give you more margin for error. I’m not sure if the twins are designed to a different standard that way, but it’s quite noticeable when you transition from singles to twins. In my experience most multi-engine training scenarios focus so exclusively on single-engine failure drills that this point is largely overlooked. Twin engine pilots rarely, if ever, actually have to accomplish an auto with both engines pulled back to idle, especially to a touchdown.

  10. Gordon,

    Good eye! One of the early comments recommended that an H/V diagram be inserted in the article. And that commenter was correct! It should have been in the article from the beginning but it accidentally got left out when the editor posted the article. The editor has now inserted a different H/V curve for the purpose of the article. The article now has the actual “generic” H/V curve out of the advisory-circular that explains Part 27 certification.

  11. Chris,

    Thanks for responding! For starters, I wish more schools would follow your autorotation regimen (zero to VNE) I applaud you for that!

    Regarding the 7000 foot reference, it can get somewhat complicated based on the helicopter. Robinson used Big Bear and I believe two other commonly used locations include Albuquerque and Roswell NM. However, when delving into Part 27 certification standards we deal with THREE subcategories under the most recent amendment number (21) and they are:

    1.) Test Conditions section

    Test conditions used are:
    “1. MGW Sea Level
    2. OGE weight lesser of:
    a. Max altitude cap
    b. 7000”

    2.) RFM Section

    This is where part 27 certification requirements dictate that the H/V diagram be listed in the performance section and not limitations (as most believe) I ask this on checkrides and maybe 3 of 10 applicants provide the correct answer.

    The standard reads:

    “H/V is performance info” and the remaining sub note reads “Max altitude for which H/V” is valid. (this subnote covers a comment Gordon made about ‘8500) It can be shown for an area above ‘7000

    3.) Remarks Section
    “If H/V is less than OGE weight H/V weight becomes limit.” It goes on to read “Applicant is encouraged to demo H/V to WAT (weight, altitude, temperature) limits.
    And the final remark goes on to say “Hover data may be shown above 7000 if H/V and IGE are demoed to 7000.

    Thanks to all for the comments.

    Matt Johnson

  12. “Sean C”

    Great point regarding the “lighter weights”.

    If memory serves me correctly [which isn’t that often 🙂 ] this is why the R22 has a “minimum gross weight of 920”.

    Matt Johnson

  13. Matt Johnson has shown with this article that he is a member of the group (students, instructors and now DPEs) who do not understand the H/V diagram.

    He states that it is developed at a 7,000 ft density altitude at maximum gross weight.

    The pertinent regulation is 14 CFR 27.87 and it states:

    § 27.87 Height-speed envelope.
    (a) If there is any combination of height and forward speed (including hover) under which a safe landing cannot be made under the applicable power failure condition in paragraph (b) of this section, a limiting height/speed envelope must be established (including all pertinent information) for that condition, throughout the ranges of –

    (1) Altitude, from standard sea level conditions to the maximum altitude capability of the rotorcraft, or 7000 feet density altitude, whichever is less; and

    (2) Weight, from the maximum weight at sea level to the weight selected by the applicant for each altitude covered by paragraph (a)(1) of this section. For helicopters, the weight at altitudes above sea level may not be less than the maximum weight or the highest weight allowing hovering out-of-ground effect, whichever is lower.

    The procedure for determining the H/V envelope is discussed in AC 27-2B, pages B-37 through B-45 and mirrors the regulation.

    He also stated that the testing is done in 5 knots or less of wind. The testing is actually done in 2 knots or less wind.

    He states that the H/V curve is based on “minimum skill level”. Nowhere in AC 27-2B does the phrase “minimum skill level” appear.

    The following statements do appear:

    “The avoid areas of the HV diagram are separated by the takeoff corridor.
    This corridor should be wide enough to consistently permit a takeoff flight path clear of
    the HV diagram using normal pilot skill.”

    “Test points with excessive gear loads, exceptional skill requirements,
    winds above permissible levels, rotor droop below approved minimum transient RPM,
    damage to the rotorcraft, excessive power, incorrect time delay, etc., cannot be

    It would seem that the diagram is based on normal pilot skill.
    He states that “The standards require the pilot’s hand to be on the cyclic and feet on pedals.”
    It is indeed a reasonable assumption that the pilot’s hands and feet would be on the controls, but it is not a stated standard.

    I am a bit concerned by the statement “getting the cyclic back immediately is absolutely critical!” It would seem to give a novice pilot the idea that he must pull the cyclic fully aft and that this must be done quickly. In reality, all that is necessary is aft cyclic to maintain the proper attitude and prevent the nose from tucking (with the attendant RPM loss).

    One other thing I would have mentioned in the article is that the H/V diagram does not apply on a normal approach. Too many times I see pilots making a high speed approach followed by a rapid deceleration and then a big pitch pull.

  14. Let me go a step farther.

    The article and the discussion appear to be based on Part 27 certification rules. Those are the issues I addressed in my first response.

    There are many helicopters which were certified under CAR 6. I have not yet researched that rule, but judging from the H/V diagrams for some older models, it would seem that CAR 6 only required testing at sea level. As such, the performance at higher altitudes is an unknown.

    I saw Matt’s posts after having posted my earlier response. In his first post, he referenced the fact that for Part 27 helicopters, the H/V diagram appears in the Performance section, not the Limitation section. I do have a flight manual for an early model Bell 47 helicopter. In this manual, the H/V diagram appears in the Limitation section. I have also been told, but have not substantiated that the same is true for the Brantly helicopter.

  15. Kris,

    Thank you for the reply. I do apologize; it is 2 knots not 5. Thanks for keeping me honest on that.

    The paraphrasing of “minimum skill level” is my writing and not a quote from an AC for FAR. As you pointed out the documents actual make reference to “normal skill level”. So an issue of semantics takes hold of the two. From my experience in evaluations, flight reviews, etc. I consider “normal” to be “minimal”. A CFI doing autorotations on a daily basis or a factory instructor pilot would be above what I consider “normal” or “minimal”.

    Regarding 7000 DA. My article was written in a general sense to not show deficiency to any particular manufacture that may do certification to only the “maximum altitude capability of the rotorcraft”. As one reader pointed out an H/V can also go beyond 7000 DA. And with that you will normally see it listed with a certain weight restriction.

    Regarding the Limitations vs. Performance section for the H/V diagram. That is good trivia. I don’t conduct exams in the two models you mentioned. Most current day trainers including the 5-7 makes/models I conduct exams in have the H/V information in the performance section.

    Thank you,

    Matt Johnson

  16. (sorry for the typo’s in above post folks – that’s what I get for typing on a small screen!)


    Thank you for the reply. I do apologize; it is 2 knots not 5. Thanks for keeping me honest on that.

    The paraphrasing of “minimum skill level” is my writing and not a quote from an AC or FAR. As you pointed out the documents actually make reference to “normal skill level”. So an issue of semantics takes hold of the two. From my experience in evaluations, flight reviews, etc. I consider “normal” to be “minimal”. A CFI doing autorotations on a daily basis or a factory instructor pilot would be above what I consider “normal” or “minimal”.

    Regarding 7000 DA. My article was written in a general sense to not show deficiency to any particular manufacture that may do certification to only the “maximum altitude capability of the rotorcraft”. As one reader pointed out an H/V can also go beyond 7000 DA. And with that you will normally see it listed with a certain weight restriction.

    Regarding the Limitations vs. Performance section for the H/V diagram. That is good trivia. I don’t conduct exams in the two models you mentioned. Most current day trainers including the 5-7 makes/models I conduct exams in have the H/V information in the performance section.

    And yes, the H/V diagram does not apply when making the approach.

    Thank you,

    Matt Johnson

  17. Matt’s article is very good, and points out a lot that is not covered in training. As someone who has demonstrated the development of the H-V curve over 100 times to test pilot and flight test engineer students, as well as a certified that the proposed H-V curve for a single engine helicopter and a couple of twins while an engieering test pilot at Transport Canada I can say with some authority that it’s a complex subject.
    Nearly all helicopters can hover OGE at maximum weight at 7,000′ density altitude, so that’s where nearly all the HV curve testing is done.
    While the standards don’t say the pilot must have hands on cyclic and feet on pedals – where else will they be? This derives from no intervention delay from failure to movement of cyclic or pedals, as opposed to a 1 second delay for collective movement following power loss.
    I defy anyone to define ‘normal pilot skill’ – the regulations state something to the effect that ‘no abnormal skill, alertness or strength’ be required for any maneuver…’.

  18. And you can read much more about autorotations in my books – Cyclic and Collective’ and ‘The Little Book of Autorotations’

  19. Hi Matt, interesting article! I do wish that people would stop using the phrase Dead Man’s Curve – I find that it takes all the fun out of autorotations! It’s rather like all the horror stories about melting jet engines that make students really nervous.

    Some machines, like the AS 355 don’t have an H/V curve at all for some circumstances, and others, like the BK 117, have one that moves it from the Performance section to the Limitations section of the flight manual once you go into high density seating. The point is moot, though, as you have to obey Performance rules anyway, so you are still more or less bound by the H/V curve. And even though the curve is not validated for approaches, the average jury* would probably expect you to obey it as it is in the book.

    *One definition of a jury – 12 people not smart enough to get out of it.

    Although, in theory, you are “safe” outside the curve, if I was in a 206 at 400 feet in the hover (the top of its low speed section), I would probably still do a vertical – it’s a lot easier than trying to go for speed and do some sort of flare at the end, and in my opinion a lot safer because that’s not stuff you practice every day. At least you can contain the mess into a smaller area. I used to be the Type Rating Examiner (DFTE) for an electricity company whose line of work was within 50 metres of 11Kv lines, that are only 20 or so feet off the ground, so I used to teach hover autos from 50-100 feet in the 206. It works very well, especially if you apply the thinking behind Dynamic Stall (as per NASA AMES) and discussed by Dennis Venturi in an ancient article in Helicopters magazine (his real name, honest!)

    OK, having said that, there is no way that I would have tried that in a 22/44 or anything else with light rotor blades, and the statement above is not to be taken as an encouragement to try it. 🙂 Autorotations of any type need to be taught by someone who is doing it every day!

    I am one of those who dislike hover autos, although I appreciate the reasoning behind them. In my opinion, all autos should be full on, otherwise you are not giving people, especially passengers the best chance. One reason is that your feet go the wrong way when you are pulling pitch, and people are more concerned with the torque spike at the end (on a 206, at least).

    Happy New Year!

Comments are closed.