Re: Maybe tracing is easier but ...
|
David J. Gall
Johannes,
Please note that all the side view drawings of the Q2/200 throughout the plans are not accurate representations of the silhouette of the actual airplane. The lower silhouette line of the side view of actual molded tail cone parts is more gently curved and the curve extends further aft than FS 120 before becoming a straight line. Your earlier CAD image looks like it has a glandular ailment in the midsection (below the main wing and behind the seatback) and I suspect it is because of over-reliance on the QAC side-view drawings. Compare:
You might do well to try to get someone to take a multitude of pictures of their airplane and run them through some photogrammetry software to develop actual lines instead of relying on the QAC pencil sketches. Also note that because of draft angle requirements for release of molded parts from molds, the sketch on page 8-2 of the plans can be a huge clue as to where the vertical tangent planes of the fuselage shells need to be located, and this translates directly to the bulkheads. For example, the vertical tangency point of the firewall is located explicitly on WL 12.1 where the upper and lower shells meet.
Reverse engineering is HARD!!! Again, it’s likely better to try to duplicate (in CAD) the physical methods used to make the actual molded parts than to rely on the sketches that were included in the plans. I don’t believe there was ever any real attempt to loft this airplane on paper and, if there was, it was held as a corporate secret. The “plans” for the fuselage are just representative sketches to aid in assembly.
David J. Gall
-----Original Message-----
Thanks Jay,
I have used these drawings as the reference. What is lacking, or maybe I am not seeing it, is how to properly constraint WL15. In particular, I never found a way to properly define the vertical position of WL15 on the bulkheads or the firewall.
The way I set up the constraints is that I took the fuselage end, at which the 1" vertical distance from WL15 is given as a dimension. I then rotated the sketch such that the firewall is vertical. This is how WL15 is approximated in my model.
I don't have any reference to the position of the fuselage split line at the firewall, so the 2.9" dimension at the firewall doesn't help.
Once that was fixed, I adjusted the fuselage skid "by hand" to match the fuselage profile to get the WL of the table surface.
Side note: the drawings seem to be distorted. I traced and scaled these drawing as well as the three view, compared and measured them. I believe the three-view profiles are properly scaled while the drawings on page 17-20 are not.
// Johannes
On 06/10/2022 15.23, Jay Scheevel wrote: > Johannes. For WL reference sketches please see pages 17-20 of the > following linked document > http://n8wq.scheevel.com/documents/Q2_Q200_Plans_Abridged_for_Scheevel > _Construction.pdf > > Cheers, > Jay >> On Oct 6, 2022, at 6:43 AM, Johannes Weissmann <jo@...> wrote: >> >> At the current stage I simply tried to reconstruct the existing shape based on the available plans and templates. >> >> Basically, I used the fuselage jig and bulkhead templates as reference and in addition traced the fuselage profiles from the three-view of the plans. The templates are to scale, the three-view needed some scaling. >> >> Based on that, I tried to make the templates and bulkheads match with as little error as possible. The main difficulty is to exactly locate WL0 and WL15 as I could not find any references. >> >> Currently, there is no limit on the degree I used to generate the curves. Some curvatures from the templates just can't be fitted with three-degree splines. >> >> >> Here is what I am not sure of: >> >> What I would *like to do* now, is to construct a design that matches as close as possible with simple mathematical forms. >> >> What I don't know because I am not an aeronautical engineer (I am a physicist) is how much I am getting into the realm of aircraft design or if the deviations from the original are small enough to not affect the resulting aircraft performance and specs. >> >> >> // Johannes >> >> >>> On 06/10/2022 14.28, David J. Gall wrote: >>> Robert, >>> If your intent is to duplicate the existing QAC shape I would caution you that, although "CAD is fun" it can also lead you astray. The state-of-the-art in the early 1980s was compass and protractor and a wooden batten as a spline; the CAD equivalents are *not* degree-three curvature-continuous NURBS curves and surfaces. Stick to degree-two conics or arcs of circles joined tangent-continuous for cross sections and nothing more than degree-three single-span curves in the longitudinal direction. If more than one degree-three span is needed longitudinally, join degree-three splines at their ends using nothing more than tangent-continuous joins. >>> If, on the other hand, your intent is to make it "better," first, define "better," then go for it! Coolio! >>> (If you're surface modeling in SolidWorks, well, there's no hope for >>> the world.... ;P) David J. Gall -----Original Message----- >>> From: main@Q-List.groups.io <main@Q-List.groups.io> On Behalf Of >>> Johannes Weissmann >>> Sent: Thursday, October 6, 2022 3:41 AM >>> To: main@Q-List.groups.io >>> Subject: Re: [Q-List] Maybe tracing is easier but ... >>> Great! >>> CAD is indeed a lot of fun. I finished the first surface model of the Q2 fuselage yesterday. Not perfect yet, but at least it is spanning the whole fuselage now. >>>> On 06/10/2022 06.27, Robert Schmid wrote: >>>> ... this CAD stuff is fun too. First version of panel test cut is >>>> done, >>>> >>>> CAD still needs some fine tuning but not bad for a start. >>>> -- >>>> Robert "TheFrisco" Schmid >>>> (408) 805-5450 >>>> >>>> www.facebook.com/TheFlyingFriscos ( >>>> https://www.facebook.com/TheFlyingFriscos ) >>>> www.theflyingfriscos.com ( http://www.theflyingfriscos.com ) >>>> >>>> Love building planes almost as much as flying. Latest completed build is "Loki", a Chinook Plus 2 bush plane. >>>> >>>> >>>> >>>> >>>> >>>> >>> -- >>> Johannes Weissmann >> >> -- >> >> Johannes Weissmann >> >> >> >> >> >> > > > > > >
--
Johannes Weissmann
|
|
|
