Friday, August 31, 2018

WRight Perspective - Article Four of Four

     This is the final article in the series discussing the 1984 Smithsonian publication The Wright Flyer: An Engineering Perspective.   This article addresses the fourth and fifth sections of the "Perspective" compilation concerning propulsion and structural design *


(The full text of the Smithsonian"Perspective" compilation pictured below can be found here.)


Publication by the Smithsonian critiqued in this and in three previous articles by Joe Bullmer

 The WRight Perspective--Article Four

 by Joe Bullmer

     As stated previously, the purpose of these articles is to address differences between information presented in the "Perspective" compilation and that in this author's book The WRight Story.   Chapters III and IV of that book contain detailed discussions of the development and testing of the Wrights’ early aircraft based exclusively on an experienced aeronautical design engineer’s interpretation of the Wright brothers' own words and records.  The discussion points in these articles have been derived mostly from information in those chapters of The WRight Story.

The WRight Story by Joe Bullmer

     The first element of the Wrights’ propulsion system addressed in this section of the "Perspective" is the propellers.  On page 80 the Wrights’ design procedure is referred to as a theory with formulas.  There were elements of theory and a number of formulas to relate some of the parameters involved.  But as explained in the first article of this series, since there are many more unknowns than knowns in propeller design, it is not a closed procedure.  Rather it is what could mathematically be termed an under constrained iteration problem. 
     To briefly recap, typically one starts with a known number of engines, their rpm and torque output, clearance restrictions, and perhaps a speed requirement.  Factors to be determined include number of blades, blade rpm, blade cross sections, angles of attack and twist, blade width or chord, taper, and resultant thrust.  One also has to account for the acceleration of incoming air ahead of the propeller which of course varies with the speed of the aircraft resulting from the thrust previously calculated.  Typically one must pick certain parameters, calculate the others, and iterate chosen parameters to a satisfactory solution.  Of course modern computer design programs with their wealth of stored data are a big help in this process.

     On pages 79 and 80, Orville’s statement in the September, 1908, issue of the Century magazine is presented, to wit “so far as we could learn the marine engineers possessed only empirical formulas, and the exact action of the screw propeller, after a century of use, was still very obscure.”  Farther along on page 80, the "Perspective" article observes that McFarland's presentation of the Wrights' procedure  is “as good a review as possible of their theory.”  It is quite possible that the reason McFarland’s compilation is incomplete is because the Wrights were not completely satisfied with the empirical and iterative nature of their own procedure and therefore did not present it in detail
     On page 81 it is claimed that early Wright propellers had problems with twisting and so “in 1905…a pie-shaped portion of the leading edge was removed resulting in a relatively constant blade width for 30 percent of the distance from the tip to the center” of the propeller.  These were known as the "clipped or "bent end" props. Twisting may have been a factor, but reducing the chord near the tips of the props no doubt unloaded the tips and improved efficiency. 
Fort Meyer, VA,, September, 1908, The Wright Flyer with "bent end propellers."

     In fact, the biggest failing of the Wrights’ original props was that their chord kept increasing all the way to the tips.  This has a similar effect to that of reversing taper on a wing.  The tip losses, and thus the induced drag, are excessive.  That’s why, since fairly early aviation, propellers have had their maximum chord somewhere before 50 percent of the blade length and taper thereafter. It is also why some historians' claims of fantastic efficiencies of Wright propellers (well in excess of the 66% claimed by the Wrights) are erroneous. 

     The article does not address the torque transmission system adequately since this was the major source of problems in 1903.  The chains used were lengths of engine timing chain and thus were much more robust than the bicycle chains assumed by many historians.  The engine ran very roughly, particularly when cold, delivering jerky torque.  So whipping and interference would still have been problems if the chains had not been enclosed in pipes.       
   Although the chains were sufficiently robust, the original sprockets, tubular propeller shafts, and bonding materials were not.  This created problems until solid steel shafts were made and proper bonding material was used to fasten the sprockets to them.

     Page 82 repeats the universal claim that “In view of the state of the art at the time, no suitable engines were available.”  This is not really true either.  Engines of twice this power-to-weight capability were in existence at that time.  More likely, something in their requirements was interpreted by the manufacturers approached as demanding a specialized engine and the Wrights were not willing to cover such a cost.  Cost was certainly a factor as they stated that they feared it was possible that crashes would destroy at least one of their engines. 

  The Wright brothers" 1903 engine (credit: The Smithsonan Institution.)

     Pages 82 to 86 contain a very comprehensive description of the design and construction of the 1903 four cylinder engine, the first built by the Wrights. The reader can’t help but be impressed with the amount of detail that had to be addressed and the number of problems that had to be solved to create this first Wright engine.  Although the crankshaft had no balance weighting, low rpm helped limit engine roughness. 

     On Pages 86 to 91 subsequent four and six cylinder engines are examined.  Unfortunately their V-8 engine was destroyed in a crash and almost no information on its design remains.  Also on page 91 the author mentions that lubrication and cooling problems arose with the Wright engines but quite correctly opines that it was in solving these problems that the Wrights showed ingenuity, not having any education or training in engine design.

     Finally on pages 92 to 94 balance and volumetric efficiency of the 1903 engine are addressed.  Not having an original or replica of the engine, efficiencies were calculated with corrections applied to data from a modern Pratt & Whitney engine. This indicated that the thermal efficiency of the Wrights' original engine appeared to have been around 25% and the volumetric efficiency about 40%. That volumetric efficiency is over half that of typical internal combustion engines at the time this "Perspective" article was written nearly a century later.

   However, the thermal efficiency is directly comparable to engines developed at the time the "Perspective" article was written. In fact in 2014, Toyota claimed to have developed an engine with a thermal efficiency of 38% which, they claim, was the highest of any mass produced internal combustion engine, most having thermal efficiencies of around 20%. So 25% efficiency for their Wright engine over a century earlier seems quite high, particularly considering that with no carburetor, no choke, and a rich fuel to air mixture, some of the fuel must have passed through the engine unburned. Considering all this, the quoted efficiencies seem quite high for an engine designed and built in 1903 by beginners.

     On page 98 the author mentions how the universal joints connecting the vertical wooden struts to the wing spars enable wing warping.  This is true and may well be the original reason the Wrights designed the joints that way.  However these joints turned out to be the critical structural design feature enabling the Wrights success as will be explained at the end of this discussion. 

Computer drawing of the Wright 1903 aircraft showing the
 force generated by the elevator and the resulting motion of the aircraft.
    Further along on page 98 a quote from Orville Wright’s 1924 letter to Alexander Klemin (footnote in MacFarland's, p. 44)
is presented that states “We originally put the elevators in front at a negative angle to produce a system of inherent stability….We found it produced inherent instability.” This unassailable quote directly contradicts information presented on page 22 in the second section of the "Perspective" and also that on page 52 in the third section.

     On the next page the Wrights’ original design specifications for their 1903 plane are quoted as calling for eight horsepower to drive a 625 pound gross weight vehicle at a cruise speed of 23 miles per hour.  It is a good thing these specifications were underestimated, otherwise, considering the ambient winds at Kitty Hawk, the vehicle would likely have been destroyed before its first takeoff.  As it turned out it was destroyed, anyway, by the winds later that day, while sitting unattended.

     Also on page 99 appears the statement that “when wind tunnel tests showed the Lilienthal coefficients to be essentially free from error…..the Wrights…..calculated the more nearly correct [Smeaton’s] coefficient of 0.0033”.  Although this statement also contradicts those appearing in previous sections of the "Perspective", it is true that the Wright’s wind tunnel verified the accuracy of Lilienthal’s lift coefficients.  But before they had built their wind tunnel, Wilbur stated in an October 6, 1901 letter to Octave Chanute that they had adopted Langley’s value of 0.0033 for Smeaton’s coefficient because it produced better correlation with their 1901 glider data.  Their wind tunnel tests commenced almost two months later in late November. 

     Pages 98 through 105 of the "Perspective" present analyses of the designs and loads of the wing ribs, spars, struts, and wires for the 1903 Flyer.  While this discussion is comprehensive, it is freely admitted that there are many gaps in what remains of the Wrights’ structural calculations and designs, and much of the analysis shown in the Loads section of his report is not necessarily what the Wrights may have used.  This discussion concludes that although the design was marginal in some respects, it was adequate and actually impressive considering the Wright’s background.

    Unfortunately this section of the "Perspective" on structures does not mention the most important advantage resulting from the structural design of Wright gliders and early Flyers.  The method of attaching the wooden struts to the wing spars is mentioned early on and the resulting compression loads in the struts are discussed.  Also mentioned is that the wires connecting the centers of the struts can be considered to cut bending moments in half thereby reducing the chance of column buckling.  However, the major advantage of the strut and wire arrangement is not mentioned.

     The attachments of the struts to the spars are effectively flexible universal joints.  Only compression and tension could be transferred through the joints.  Bending could not.  The crossed guy wires were in tension and thus pre-loaded the struts in mild compression.  Photos of crashed or hard landings indicate that the easily replaced wires were the first things to fail, usually leaving the wooden struts intact.  This was aided by the wires connecting the centers of the struts causing adjacent struts to share any column bending loads resulting from excessive compression. 
     Consequently the vehicles could survive hard landings and moderate crash loads with only minor repairs to the biplane structure, often just replacing some wires or “sistering” a split skid with a spruce stick and some twine.  Not only that, but the universal joints were achieved with simple metal hooks and loops which were impervious to sand or dirt.

  The net result was vehicles that could survive extended usage and numerous mishaps with only minor repairs if any.  This was the all important feature allowing the Wrights, over relatively short testing periods, to make well over 1,000 glider flights and over 150 powered flights to determine design improvements and perfect their flying skills.  It is not an exaggeration to say that without this flexible and quickly repairable structural design the Wrights would probably not have been able to make manned, powered, controlled flights by the end of 1905.**
     This concludes the fourth and final article in the four part series discussing The Wright Flyer, An Engineering Perspective.  It is hoped these articles will be treated in the spirit intended, merely to make the historical record of the Wrights’ work as accurate as possible.  Further detail on the points made in these articles, as well as references to sources, can be found in the author’s book The WRight Story.

Joe Bullmer, above, has a Master's degree plus advanced studies in Aeronautical Engineering. His first contribution to the"Truth in Aviation History" series of articles is "Joe Bullmer Rebuttal to Tom Crouch in the"Huffington Post." about the claimed fourth flight picture of the Wrights in 1903.     

All of the pictures and most of the links in this essay were selected and added by the founding editor of "Truth in Aviation History."

* Links are provided below to the first three articles of Joe Bullmer's critique of the Smithsonian publication: "The Wright Flyer: An Engineering Perspective":
1. Article One,  
2. Article Two, and 
3. Article Three

**Editor's note: There are no legitimate records of witnessed, unassisted take offs by the Wrights until years after the 1905 date.


Sunday, July 8, 2018

WRight Perspective - Article Three of Four

        The  WRight  Perspective – Article  Three
By Joe Bullmer

 The two illustrations including captions ( above) are from the publication The Wright Flyer, an Engineering Perspective

   This is the third article in a series discussing the Smithsonian compilation document The Wright Flyer, An Engineering Perspective, cover pictured below.*

    [The two previous] articles have addressed the section discussing the Wrights as aeronautical engineers and the section on aerodynamics, stability, and control.  This article discusses the third section, titled Longitudinal Dynamics of the Wright Brothers’ Early Flyers. 

  The purpose of these articles is to address differences that have been pointed out concerning information presented in this author's book The WRight Story and that presented in the Perspective.  The purpose of The WRight Story is to record an accurate description of the Wright brothers’ work.  Original documents supporting the following comments are referenced in that book.

     At the start of this section, page 45 of the Perspective, the author [Frederick J. Hooven] presents an entertaining description of his close relationship with Orville Wright from 1925 until Orville’s death in 1948.  Unfortunately, throughout these years Orville apparently allowed the author to believe that Otto Lilienthal's lift data were found to be wrong.  On page 48 the author states that at “The end of the 1901 season…..having found Lilienthal’s data to be mistaken….. they realized that they would have to develop their own aerodynamic information…”  This contradicted a 1902 letter Wilbur wrote to Octave Chanute saying that Lilienthal’s data “were as accurate as is possible."

Otto Lilienthal (1848 - 1896) and one of his magnificent gliders
   Even more amazing in view of his lengthy and close relationship with Orville, the Perspective author was apparently totally unaware of the instability of early Wright aircraft until he read Charles Gibbs-Smiths 1966 book The Invention of the Aeroplane. 

    On page 46 of the Perspective the author claims that book “made the first mention I had ever seen of the longitudinal instability of the Wright machines.”  So evidently Orville never mentioned it.  In fact, the author claims to have not believed it until he commenced two-dimensional computer simulations in 1978.

  On page 47, Edward Huffaker's note from July 29, 1901 concerning the Wrights’ glider is quoted stating that “The equilibrium is not satisfactory and the Wrights think of making radical changes, placing the rudder in the rear, or rebuilding the machine”.  Interestingly, the surviving notes of the Wrights as presented in McFarland's compilation** mention nothing about considering an aft elevator after mid-October of 1900. 

   Page 49 notes that the Wrights tried to reduce the pitch instabilities of their vehicles by altering center of gravity (C.G.) locations.  They first moved the C.G. of their 1904 machine from the 29% chord point farther aft to the 32% point.  Finding this worsened the pitching problem, they reversed the C.G. location to the 23% point by adding 70 pounds of iron to the canard supports.  Obviously, although they knew the location of the C.G. was important, the Wrights didn’t yet understand the control implications of the C.G. location relative to the center of lift.  Actually, the Wrights did not develop a way of determining the locations of the centers of lift of their vehicles.

    Page 50 begins with the statement that by 1905 the Wrights had eliminated the pitching problem with their aircraft.  “From August 24th [1905] onward…..there was no longer the tendency to undulate [in pitch] and there were far fewer crashes”.  This contradicts movies taken in 1909 from their aircraft in flight clearly showing constant rapid movements of the elevators to cope with pitch instability.  

[A 1909 inflight movie of a Wright plane shown above. It doesn't take a rocket scientist
 to note the pitch instability. Ed.]

Image result for glenn curtiss
Glenn Hammond Curtiss (1878 - 1930), the pioneer who got it right..
  This statement also conflicts with another on page 51 that implies that, if they’d had more time, the Wrights might have put their elevators in the back.  The interesting observation is made that the Wrights kept their unstable canard design after 1905 because “They had quite enough to do to build more machines and prepare for public flying without trying to develop a radically different machine for 1908”.  Obviously one of their real fears was that they would fall far behind others that were building flying machines, especially Glenn Curtiss  in the U.S. and numerous others in Europe.  Another may have been that the canard being an obvious feature, abandoning it would have made their patent much
less enforceable. Either of these outcomes would have jeopardized their chances of cashing in on their work.

    On the next page appears the popular statement that “the Wrights conceived the airplane from the very first as a craft that, like the bicycle, depended upon its rider to maintain its equilibrium.” As discussed in the previous article in this series, this is not true. In a legal deposition written in 1920, Orville wrote that they originally thought they would have a stable machine because the data they had on hand showed the center of lift to move exactly opposite of the way it actually did with changes in angle of attack.

   On page 52 it is also mentioned that the wind at Kitty Hawk when the Wrights flew in 1903 averaged about 20 mph.  This is curious since other sources, including Orville Wright and the National Weather Service, claimed it to be 25 to 27 mph.

    On that same page it is pointed out that the Wrights suffered almost two dozen destructive crashes in 1904 and 1905 learning how to turn their aircraft.  This clearly contradicts the first section of the Perspective in which the statement is made that the 1902 glider (and by inference the 1903 Flyer) was capable of making “smooth banked turns.”

   From page 51 to 58 some of the theory and results of two dimensional computer flight simulations of the 1903, ‘04, and ‘05 Wright Flyers are presented.  These simulations do not attempt to represent any actual flights by the Wright aircraft.  They were done to determine if the vehicles actually were longitudinally unstable.
   Since the simulated vehicles were indeed longitudinally highly unstable, a simulation of a pilot’s control actions had to be included to avoid immediate crashes.  Pages 53 and 54 reveal that the average pilot reaction times used for pitch control in these simulations were 0.04 seconds for the ’03 and ’04 aircraft and 0.05 to 0.07 for the ’05 and ’07 aircraft.  Any more and they would crash.  One wonders how realistic a pilot reaction time of 1/25th of a second is, particularly since there is no requirement for lateral control in these two dimensional simulations.  Also, allowing slower pilot reactions for the more stable ’05 and subsequent aircraft indicates that reaction times were set by the requirements of the vehicles rather than the improving capabilities of the pilots.

    Page 70 points out that wind tunnel tests of accurate models of the 1903 Flyer showed the drag of the complete vehicle to be higher than the Wrights’ composite estimate.  But it then says that horsepower of the engine was correspondingly higher than claimed because of ingesting cooler air near the sea and better cooling with the cooler air.  It is further claimed that, due to proximity to the ground, induced drag would be lower than measured in the tunnels.  And finally, with no explanation it is stated that propeller efficiency was better than the Wrights claimed.

   The validity of the simulation results could be questioned because of the assumptions just described regarding power, drag, and the propeller.  But perhaps the most damaging assumption affecting the usefulness of the results is the two dimensional nature of the simulation.  The title of the article does state that it is limited to longitudinal dynamics, but how useful is that if the pilot doesn’t have to concern himself with roll control of a vehicle that was also laterally unstable because of its anhedral?  In such a vehicle a substantial percentage of the pilot’s attention must be devoted to roll control, certainly at least 20 percent.  This could well distract the pilot from pitch control for a second or more, allowing the aircraft to crash from pitch instability.  In fact, in a September 20, 1902, entry in his diary, Orville admitted that he had crashed for exactly that reason

  The final five pages of the section is a line listing of the simulation program. 

   The next and final article in this series will discuss the sections of the Perspective concerning propulsion and structures. 

   *The document is available online in various formats here. Because we frequently refer to specific sections and page numbers, we recommend downloading the PDF version, available here.

  **Marvin McFarland's complete compilation is online. We have linked you to Volume I of the two volumes.

Joe Bullmer, above, has a Master's degree plus advanced studies in Aeronautical Engineering. His first contribution to the"Truth in Aviation History" series of articles is "Joe Bullmer Rebuttal to Tom Crouch in the"Huffington Post." about the claimed fourth flight picture of the Wrights in 1903.     

All of the pictures and most of the links in this essay were selected and added by the founding editor of "Truth in Aviation History."