Jump to content

EAS100 New Design so far


EAS100

Recommended Posts

OK, I am just getting started here, and have a LONG ways to go. But, the basic configuration is now apparent.

 

This is a plane I designed back in 1984, as a requirement for getting a degree in Aerospace Engineering. I call it the EAS100.

 

It is a three surface configuration, with pusher propulsion. Much like the Piaggio P180 Avanti. My design is much larger however, made to transport up to 80 people. Another difference is the use of counter-rotating turbo-propfans. This is sort of an attempt to model the General Electric UDF prototype, and test bed was actually flown.

 

Wing airfoil is the Grumman K-3, extruded from an accurate template. It is transonic capable. The canard and tails are a supercritical NACA 11% thick, also generated from extruded template.

 

I'm still learning the ins and outs of FSDS. However, I have translated this design into CAD space, and now I'm looking at some virtual wind tunnel programs to see how that works out. My plan is to take either DATCOM, or some vortex lattice method, and estimate stability derivatives to begin work on the .air files. My old equations and charts show it would fly per the original RFP. Again, still hundreds of hours of work yet to do, but thought you might like to see something that I have dreamed about doing for over 30 years now.

EAS100 021716A (Medium).jpg

Link to comment
Share on other sites

I might have jumped the shark (so to speak) here, but in my quest for a more complete knowledge and understanding, I am going down a new path. Been hammering away many hours lately, to get an understanding of computational fluid dynamics (CFD). I've really not studied up on aerodynamics in over 30 years now, and the field of course has advanced by leaps and bounds.

 

However, I started with the fuselage only, exported to a CAD file. Used Blender to repair and seal mesh geometry, and set up a flow field. Then, exported this to enGrid and generated mesh elements over the fuselage, a boundary layer, and mesh elements for the chamber flow field. Never having used either of those programs before, it has been an El Capitan learning curve (very steep).

 

enGrid can export to openFOAM, a open source fluid dynamics program. There I am stuck for the time being. Also 30 years behind on programming. My last experiences with that were Fortran with punch cards, and Pascal on a command line only CRT. I know nothing about C modules, Python scripts, etc. More study there to even get started.

 

Here is a screen shot of the enGrid screen, with fuselage and flow field mesh. The red geometric blocks are a cutaway of the volume elements in the flow field. This took many hours of study, trial and error plodding, and dumb luck. But I am learning many new things.

 

This is just another facet in my quest to become a more skilled aircraft designer.

EAS100 FlowField (Medium).jpg

Link to comment
Share on other sites

An update on what I have been doing. After researching a few things, I decided that OpenFOAM was not the best solution for me. I found another program called SU2, that was more in line with aerodynamics. After much difficulty trying to get it to run on my Windows desktop, I decided to re-task an old Vaio Pentium 4 system I had, with Linux Ubuntu operating system. A whole new world there, as this was not in my skill set, but I hammered through it.

 

Finally got the Linux system operational, and took a few days to get to where I could use it. Loaded SU2 via a remote control script from a kind soul at M.I.T., and figured out right quick that doing my 3D mesh would take a VERY long time. I decided to alter a tutorial file set, and generate a 2D aerodynamic solution for a simple NACA0012 wing section. As I gain more experience and knowledge here, I will eventually get a full set of flight dynamic coefficients to plug into either an .air file, or something else (looking at X-Plane very closely now).

 

Now the results from the SU2 CFD program are just plot files generated, and they have to be visualized in some other program. Being retired and on budget, I found Paraview to be the peach for this. All these programs I am using, are quite complex. enGrid, Blender, SU2, and Paraview are all public domain and free for use. All have very steep learning curves. Here are a couple screen shots from the Linux Ubuntu file computations, and then a quick visualization of a pressure field over a 2D wing section in Paraview from that computation. One thing for sure, I will need more computing power for the SU2 system, and already looking at some used i5 and i7 systems, where I can run multiple cores.

Screenshot (Medium).jpgPviewNACA0012 (Small).jpg

Link to comment
Share on other sites

After many hours of study, software thrashing, mistake after mistake, I finally got a virtual wind tunnel solution on just my fuselage. At one point, had to go clear back to FSDS, rebuild the fuselage, export to CAD and realign axis system, then go to Blender and tweak the mesh. Then, into enGrid to set up wind tunnel chamber, and stumble like a drunk gorilla in a china shop with boundary layers, inviscid walls, prismatic elements, and well...you get the idea. Struggling to take 1984 knowledge, and update to 2016 computational advances. First shot here is fuselage in Mach 0.75 freestream. Second shot is theatrics, with virtual air particles streaming across fuselage. The hard data stuff is in the form of charts, spreadsheets, and other forms that I can plug into a flight dynamics file. Surprisingly, this only took 10 minutes of solid crunch time on a Linux Ubuntu Pentium 4. Resolution of the finite elements is about average I guess.

 

The only software that cost me money, was FSDS. The rest has been open-source free to the public. A few more runs with fuselage at different angles of attack, and side-slip conditions. I should be ready to tackle a half-wing section in another month or so...

FUSMACHp75 (Medium).jpgFUSMACHffp75 (Medium).jpg

Link to comment
Share on other sites

This is a thread I'll follow very closely!

 

As an active AE, I've obviously looked at having my own OpenFOAM wind tunnel (as the WAF of building a water tunnel in the living room reached staggering new lows, even compared to the offroad bike on the porch and the fighter cockpit in the library, and the one-week-later suggestion of building an aquarium for streamliving fish was met with suspicion). Unfortunately, as you noted, the learning curve and time investment is steep. Between fulltime work and toddler, it's never gotten anywhere. Thanks for posting your endeavours. I'll live through you here pending my own promotion to full-time tinkerer in some 25 odd years. :)

 

Cheers,

Fred

Link to comment
Share on other sites

Thank you Fred. I'm taking a short break here now, to get things around the house squared away. I've been neglecting some other things in life, in my focus to get my virtual wind tunnel operational. It's all good though.

 

Yes, having a full time job to hold down and family to support would make it quite difficult to put in the hours of concentration and study that it required to make this thing come alive. Retirement has been the greatest thing that I have ever experienced, as it is like coming out of a cage. I intend to make the most of it.

 

I sort of think now, things happened the way they did for a reason. All I had in 1984 was a TI-59 calculator, and a very primitive Timex-Sinclair 2068 computer. I was programming the 2068 with a BASIC program I wrote myself, that solved control system polynomials with a Bairstow method algorithm. Amazingly, I sold that computer about 8 years ago, for more than I paid for it on eBay. (Kept the box and manuals and everything from original though). Now, I have some tools that were far beyond my imagination back then.

 

My next endeavor will be the wing in the SU2 CFD. It will be really interesting to see what the lift and drag forces will be. I'll update here when I get some graphics and numbers.

 

Never give up hope. It's the main thing that kept me going all these years.

Link to comment
Share on other sites

Things went a little faster than I thought. I am not so lost now with some of these software packages, and maybe I can do 3% of what their function capabilities are, versus 1% a few days ago. I had to completely restructure the wing, and believe it or not, I made a half wing mesh completely in FSDS. Because of the wing profile, I had to do point by point polygon work on the end caps of the wing. My patience was rewarded with a good STL export, that was watertight. Then, in Blender, I generated a full wing and the fluid domain. Then, onto SU2 and the computation took about an hour with 200 iterations. This Paraview is pressure field across wing. Plenty of lift, even at zero angle of attack. C-lift at 150 iteration around 0.455. L/D at 14.6 Oh, this was Mach=0.5, and flow on wing top was max Mach 0.68. I am running another now at Mach=0.75 to see if flow goes transonic. I want to see how well this Grumman K-3 airfoil wing does. Bit by bit, I'm getting there.

WingPressure030416 (Medium).jpg

Link to comment
Share on other sites

Made some interesting progress on the project. I have been running multiple wing virtual wind tunnel scenarios. Various Mach at various AoA. This will give me a good idea of the Cl, Cd, and Cm vs. AoA curves in different flight regimes. I let those simulations run on another computer, while I work the design from my main system.

 

I wanted to progress, and finally try to combine fuselage and wing in the fluid dynamics space. I wanted variable meshing on each aircraft component, and while FSDS can do a lot of things, it cannot import STL mesh files. Blender to the rescue. I was able to export separate finite element mesh configurations from enGrid back to binary STL format, and combine in Blender. Then, I was able to take my backdrop image for general aircraft assembly and make it visible in Blender (just like FSDS). Very nice, as it is has many more adjustments than FSDS does. Blender seems to be the great go-between from FSDS to fluid dynamics analysis in SU2. This gives me a new range of powers to precisely tune the aircraft configuration. enGrid of course, is essential to generating the boundary layers, fluid cell, and volume elements needed by SU2.

 

One major thing, is the axis system used in FSDS, is completely wrong for SU2. Again, Blender makes quick work of reorientation from one axis system to another. Quite amazing, that Blender is free open source software. Just amazing.

FuseWingBlend (Medium).jpg

Link to comment
Share on other sites

I hit a brick wall a couple of days ago. When I first tried to mesh the wing/fuselage combination in enGrid, I got nothing but crashing grief (literally). Finally had to learn rudiments of Boolean logic operators between two objects, and then search for a finite meshing program that was better than enGrid.

 

Even though Blender reported a watertight mesh with no open manifolds, edges, or what not, it still crashed in enGrid. FreeCad would destroy the entire shape "repairing" what it considered to be mesh faults.

 

Finally found a wonderful program called "Graphite" (also free and open source), that did quick work of repairing mesh without damage, and then generated a nice even set of finite elements over a combination of components.

 

So, long story short, Boolean logic joins wing/fuselage as one structural mesh in Blender. Graphite repairs structural mesh beyond Blender, and generates a new even FE mesh over the repaired structural mesh. enGrid likes this new aircraft mesh, and then is only used to generate fluid cell volumetric mesh. Export system to SU2. SU2 runs the system, and I have a new result Paraview.

 

These screenshots are at free stream Mach =0.5, and most evident is acceleration at wing/fuselage interface. Need to do some radius blending work there for sure. Next, I will attach canard, and see what happens.

FWMp5 (Medium).jpgFWMp5i2 (Medium).jpg

Link to comment
Share on other sites

I just download freeware and payware - thousands of flight sim enthusiasts, myself included, cannot possibly fathom the number of hours, or the amount of effort takes to create a realistic simulation of even a real-world aircraft, not including the liveries, or original panels.

 

So I sure applaud guys like you...

Link to comment
Share on other sites

Thank you UPHILL3. Really, this is a labor of love. I just get a ton of satisfaction from this endeavor. I spent over 30 years, working at something that gave me ZERO personal satisfaction, and all I could dream about was someday I would be free to do as I pleased. That day is here now.

 

At any rate, I took delivery today of a used Lenovo i5 barebones computer. It did have 8GB of RAM, and 500GB HD for a bargain basement price. eBay is my friend sometimes. I spent the entire day erasing Windows from it, installing 64-bit Linux Ubuntu, and finally a compiled SU2 suite. This computer is intended for just fluid dynamics, and perhaps in the future, structural simulations using finite elements.

 

I finally got a mesh-able fuselage/wing/canard configuration made. I had no end of difficulties with Boolean operators, and getting surfaces to mesh. But hours of hammering away, and I got some results. I will attach image of last simulation run, Mach=0.5 and AoA=0. Here is the interesting thing. I ran same simulation three times, in three different computer configurations. 200 iterations each run.

 

The old Pentium 4 (one core) took 5794 seconds. The i5 on one core took 1296 seconds, and then I hit the throttle on i5 all four cores, and took 481 seconds. So I can run different scenarios a lot faster now. I have only just started learning the capabilities of these software packages, and as time goes by, I am sure I will expand the skill set.

FWC Mp5 (Medium).jpg

Link to comment
Share on other sites

Here is the same view with the elements visualized. That Graphite software is quite incredible. I was able to generate 10K, 15K, and 20K meshing grids in just seconds. But the imported model has to be PERFECT, and that is easier said than done. This is the 15K element count. FWCmMp5 (Medium).jpg
Link to comment
Share on other sites

I finally have some early results of flight coefficients of the main flying combination. Empennage was added, in steps, but I decided to run the virtual wind tunnel with the entire T-tail. I was really interested in CL and CD curves with all flying surfaces.

 

With the new (used) i5 system, I did a few tweaks and was getting an iteration every 1.2 seconds, so it was about 4 minutes per run of 200 iterations. Much better than waiting over an hour and a half per run. With all four cores running at max, the computer cooling fans are on full speed the whole time.

 

Here is the full configuration of flying surfaces, Mach=0.5 at 5 degree AoA. I made 24 runs at various angles of attack from -21 to +21 degrees. There is no twist, nor wing-fuselage incidence on these numbers. L/D maxes at 15 at +2 degrees, which is not great, but given the wing loading, it is acceptable to me. I might enlarge the wing a small amount, but then of course induced drag goes up too. A lot of streamlining details are needed as well.

 

I'll add more virtual runs at extreme angles of attack, and then start in with sideslip conditions to evaluate roll moments and side forces. Initial chart of CD and CL vs. alpha here is more for a quick look see of where I'm at.

Mp5AoA5031316 (Medium).jpgEAS100 Coeffs031316 (Medium).jpg

Link to comment
Share on other sites

Update 3/17/16: Several things of note in my progress lately.

 

1. I enabled the Linux i5 SU2 system to be remote controlled by the Windows 7 i7 computer, with shared folders. I got tired of going back and forth from machine to machine with a thumb drive.

 

2. I downloaded a full science Python (SciPy) package to the Linux system, so I could start running Navier-Stokes fluid dynamics simulations. The Euler simulations were good, but they really do not take skin friction into account, only pressure. I fully implemented a Reynolds Averaged Navier-Stokes (RANS) wind tunnel simulation on the plane. I was not happy with what I saw. Lift and drag were not what I wanted.

 

3. I started from square one, and made new components. So far fuselage and wing. I constructed a fuselage-wing interface that was more like you would see in the real world. A belly pan to stow landing gear as it were. I had to construct wing root shapes, that would connect fuselage and wing, and it turned out pretty good after long battles.

 

4. Boolean logic operators on FSDS.....nothing but grief. Error this and error that. I built components in FSDS, and exported as .x, then convert to binary .stl. This way dimensions were consistent from one domain to the other. The wing wizard on FSDS completely messes up dimensions it seems, so I built that from scratch again.

 

5. STL components were imported into Blender. Boolean logic there is quite tricky, but you have absolute control. Normals have to be just so, and I did a bit of vertices stitching to get water tight mesh frameworks, but at least no fatal errors if I did my part right.

 

6. Finally, after quite a long session, I had one contiguous, single shell mesh. This was exported back as stl into Graphite. Full control over finite element meshing there. Export again, and then import into MeshLab, where there are several refinement controls. Mainly shape constrained Laplacian smoothing transforms. Automatic streamlining of some rough bits. Export as stl again, and back into Blender to generate the fluid cell domain. Export the system as .begc file into enGrid, and make sure to check all geometry again with every single tool there. Generate boundary layer and volumetric mesh. After that, an SU2 file can be generated, and ready for the virtual windtunnel.

 

First screenshot from Blender. (This has animation capability too, along with a thousands things that do not seem doable in FSDS). The backdrops can be micro-adjusted, even for rotation if your .bmp file is not perfectly orthogonal.

 

Second screenshot from showing cutaway of shell. Those wing root blends were quite tough for me. Any stray bits or pieces, bad normals, or anything less than robust mesh geometry will cause enGrid to crash.

 

This process is complex, but yields results with mostly open-source software. Again, the only thing I paid for was FSDS (the Boolean logic operators nightmare).

Blender001 (Medium).jpgShellCut (Medium).jpg

Link to comment
Share on other sites

I am immensely pleased right now. After spending this day, reconfiguring the wing again, I found a major issue. If a swept wing is designed with the airfoil sections parallel with the flow, you really lose the effect of the sweep. Wing and cylinder extrusions in FSDS do this, and it really negates the purpose of the sweep. I also added a 2 degree washout twist, and only rotated the wing half to sweep after it had been built.

 

Also learned new tricks in Blender to do Boolean operations more effectively. I have to keep a clipboard by the computers to take notes on each complex step I took. With proper belly pan for landing gear, and wing roots blended in, it got a little trying. In a flight simulator, all this would not really matter, as you just plug in the numbers into the .air file, or whatever to mimic the behavior. But where do those numbers come from in the first place? Wind tunnels and actual flight data. Here I have to generate a real mesh-able shape that a wind tunnel would see, and then run it. Anyway, my first wing/fuselage combo on 3/9/16 had a 15.6 Cl/Cd at Mach 0.5 and zero AoA. Just a short while ago, I ran a new combo at same conditions, and got a 27.3 Cl/Cd!! These are Euler equation numbers, but bodes extremely well for a Navier-Stokes run improvement.

 

Screenshot 1 is revised wing construction to take advantage of sweep and twist.

 

Screenshot 2 is the Mach plot of todays run. Pressure is much more even, and well distributed across the wing. In short, lift is substantially up, and drag is down with the recent modifications. Washout twist really helps.

 

Just pleased as punch. Now to tackle the rest of the configuration again.

 

SweepAngle (Medium).jpgEAS105a (Medium).jpg

Link to comment
Share on other sites

  • 1 month later...

Been over a month since I last posted on this project. I've been neck deep in getting spring garden prepared, and repairing various internal combustion engines that all seem to have gone balky at the same time. But all this is not keeping me from my main interest here.

 

I got completely disgusted with enGrid as a meshing program. I spent more time struggling with it, and having it crash on me. The worst was when I added the nacelles to the configuration. As such, I learned the rudiments of a new meshing program (also free and open source) called Gmsh. It gave me much more information, but took a long time to get my feet on the ground with it. A lot more precise and deliberate steps are required, but once learned, it makes life a lot easier. I no longer have to rely on Blender to generate fluid cells, and now can simply substitute a .stl file name in the program execution text file to quickly get a new system ready for virtual wind tunnel use. Major time savings on that.

 

In addition, I have gone through extensive weight and balance calculations. Back in 1984, I basically used what are called "Class 1" methods. Now, I am using a more accurate and precise "Class 2" methodology. Further, I can research on the internet, and find weights of various components such as engines, APU, and other engineered items. I reduced weight of several items, especially electronics and navigation from estimates 32 years ago. Carbon fiber figured into several items. A spreadsheet was made just for doing quick weight and balance changes, which I would have paid a lot for back in '84. Now, I have OpenOffice (a free open source MS Office substitute).

 

At any rate, I am inching closer to generating flight dynamics equations using Digital Datcom. Comes to find, Datcom cannot handle three-surface configurations at the outset, but it does allow substitution of experimental data for the very coefficients generated by the virtual wind tunnel runs. That is, Cl, Cd, Cmx, Cmy, and Cmz. Hence, I have been hammering away further into computational fluid dynamics. My skill set continues to expand and grow each day, as I learn and relearn the art and science of aircraft design.

 

Screen shots here are first wind tunnel run on full configuration. A GMSH generated finite element fluid cell field on a 20K resolution FE aircraft, and finally a shot of weight and balance sheet. I added a fuel tank forward of the wing and landing gear stowage of a "half moon" cylindrical section. The intention is to have this function as a ballast to adjust c.g. in relation of aerodynamic center. Of course "virtual pumps" can transfer fuel from wings to forward tank back and forth.

 

At some point, I will do flight dynamics after I get Datcom pegged down enough to use it.

 

Lastly, and I sort of dread this, it will be back to design and try to figure out textures to do the paint scheme, animations for flaps, landing gear, and all that. Probably be down the road quite a ways at this point. The outdoors beckons with arrival of spring and warm weather.

 

[ATTACH=CONFIG]189093[/ATTACH][ATTACH=CONFIG]189094[/ATTACH][ATTACH=CONFIG]189095[/ATTACH]

Link to comment
Share on other sites

  • 3 weeks later...

Dynamics: Quite a bit of progress since last post. I went ahead and purchased a piece of software that I was convinced would help me out. I will give a plug for DATCOM Plus Pro here. It makes calculating effects of flaps, ailerons, elevators, and propulsion on a configuration quite fast and easy. I've also been using a freebie called XFOIL, in conjunction with my SU2 wind tunnel to further analyze lift, drag, and pitching moment numbers.

 

Further, DATCOM Plus Pro generates a .xml file for use in JSBSim, to analyze dynamic behavior in flight. Not quite there yet, as I needed extensive analysis of C.G. in 3D space, and then the resulting moments of inertia. To that end, I created a spreadsheet that calculates all these numbers, based on component weights and positions in 3D space. Since simulators typically require moments of inertia, I wanted to have accurate numbers for most accurate simulation.

 

I redesigned the fuel system, creating 5 separate fuel tanks. 4 of which are integral to the wings, and one which is a bladder type forward of the landing gear bay in the belly. With this, I can adjust CG in flight if so desired. I used Roskam's Class 2 weight estimation methods, and also his moment of inertia equations. The spreadsheet even gives cross product moments Ixy, Iyz, and Ixz. DATCOM Plus Pro uses those numbers to generate the dynamics model for JSBSim.

 

Hopefully, by end of the month, I will have my first full dynamic model with all the relevant coefficients.

CG&MOI (Medium).jpg

Link to comment
Share on other sites

Thanks for keeping us updated on your progress. It does appear that you are having fun at least! :)

Bill Leaming http://smileys.sur-la-toile.com/repository/Combat/0054.gif

Gauge Programming - 3d Modeling Military Visualizations

Flightsim.com Panels & Gauges Forum Moderator

Flightsim Rig: Intel Core i7-2600K - 8GB DDR3 1333 - EVGA GTX770 4GB - Win7 64bit Home Premium

Development Rig1: Intel Core i7-3770k - 16GB DDR3 - Dual Radeon HD7770 SLI 1GB - Win7 64bit Professional

Development Rig2: Intel Core i7-860 - 8GB DDR3 Corsair - GeForce GTS240 1GB - Win7 64bit Home Premium

NOTE: Unless explicitly stated in the post, everything written by my hand is MY opinion. I do NOT speak for any company, real or imagined...

Link to comment
Share on other sites

  • 3 weeks later...

OK, I got really burned out on flight dynamics for a while. Started doing some flight surface superposition in DATCOM+. Got really deep in the mire, as I started making new wind tunnel runs on various configurations to feed experimental data coefficients into the DATCOM+ program. I had to give my old calculus skills a dusting off. Ended up with a fourth degree polynomial to fit the wind tunnel points on CL-alpha, and then took a first derivative to get the third degree polynomial to figure out the curve slopes at individual angle-of-attack points. I gathered I was into some not-to-frequently ventured territory on DATCOM+, as most users just input airfoil coordinates into the program, and let it do the lift calculations. Comes to find that DATCOM totally freaks out on thick supercritical airfoils, so I had to use the experimental data overrides in it. DATCOM was giving negative lift numbers even up to +15 degrees AoA, so it was obviously not made for that, and it does say so in the literature. However, it digested the symmetrical airfoils OK, and I got comparable results to wind tunnel and XFOIL analysis there.

 

I got to sizing ailerons and elevators, and after a many runs of DATCOM, I finally got the plane to trim out at cruise, approach, and landing. I still have problems with some stability modes, and as yet have not figured out how to implement the canard control system in relation to the tail elevator. I need to let that one stew a while in the cranium.

 

So you see, my brain was toasted after a couple weeks of that.

 

So, I ventured into the graphics territory and integration with FSX. My first little bit, was doing an animation of landing gear using Blender. I think I have just about totally abandoned FSDS for this sort of thing. The more I use Blender, the more I see the superior control it has. I rebuilt the entire plane using separate components again, now that I did not have to create a single smooth watertight shell for wind tunnel analysis. More learning on key frames, and then into materials and textures. Good grief, I will be ready to go back to flight dynamics after a bit more of this.

 

I rigged the plane with the .cfg and .air files from the Bombardier CRJ-700 since it is about the closest thing to my plane there is. Tweaked the contact points a bit, and got my plane on the runway at Amarillo, TX and took it for a spin over the Texas Panhandle. Gear goes up and down very nice, but no doors or details yet. Just getting my feet wet here for now.

 

Lastly, fought the texture orientation on Blender to FSX for a good while. Finally got my personal logo on the tail in correct orientation. Had to flip and rotate in GIMP a few times to get it to match up OK. Oh yea, that reminds me, the axis system had to be re-oriented completely once again, which was another fine mess after putting all the animation elements for the gear in.

 

At any rate, here is my plane in FSX with simple materials and simple model, with operational landing gear. Still a long ways to go, before this gets to where I want it, but I have the time to do it now.

 

And yes, I am having quite a bit of fun with all this.

AtTakeoff (Medium).jpgSkimmer (Medium).jpg

Link to comment
Share on other sites

  • 3 months later...
This project will be put on hold for quite a while. One of my parents has chosen hospice care at her home, and I have been travelling back and forth from my home in Tennessee to theirs, a couple of states away. I have no idea of when I will be able to pick back up on this.
Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...