A minor release containing various improvements that have accumulated during the remainder of 2012. This update brings a more realistic solar system model derived from orbital elements, adaptive step size in numerical integration and performance improvements. The model file now includes Ranger Block I/II and Surveyor Block I, Falcon 9 (LEO) and a very sketchy Falcon 9 Block 2.
ORT 3.0 is now available for download. It is probably the largest update in the history of ORT to date and is best summarized as "proper support for the third dimension". ORT Live has gotten quite a boost through integration of JME3 for rendering, which of course caused ORT to put on some serious weight, now measuring in at a 12 megs download. The modeller has received rudimentary support for configuring the launch trajectory and aligning the trajectory view to the trajectory plane. This release also brings extended documentation with 3 PDFs total, as well as new and improved vehicle models. Development now focuses on more advanced GNC commands and proper simulation of the solar system to allow interplanetary travel. Model-wise the focus will lie on the Falcon 9 which is due to launch to the ISS end of April.
Work on ORT 3.0 is progressing well. The current head contains a simulation configuration dialog which allows setting up launches from any point on a celestial body defined in the model file, specified by means of spherical coordinates (longitude/latitude/altitude).
Using this feature I ran a simulation of a standard V-2 launched from the lunar equator. Configuring the pitch as start = 9 s, end = 59 s and target = 147° will put the rocket in a lunar orbit with an apoapsis of 163 km.
This release brings the Titan 2 model, shortcuts for varying simulation speed in the LiveApp, a new GNC command for throttling engines (put to use in the V-2 model), the capability of loading and saving different model files in the ORT modeller and modelling of celestial bodies and atmospheres. Also included is another small fix related to atmospheric drag.
I implemented Cd curves and further refined the Qu8k model. Enough has come together to warrant a new major, so I put up release 2.0.0. Included is the revised modeller manual, which I have migrated to LaTeX because fiddling around with the images in LibreOffice took a disproportionate amount of time.
I've updated the ORT screen shots to show off some new features. ORT now flies engines with thrust curves, and I've updated the Qu8k model to take advantage of this. I'm quite pleased with how close to the actual performance the simulation already gets. I also did some UI work, so now the various panes in ORT Modeller and ORT Live can be resized. Next up will be Cd curves for nosecones, which should help with accuracy of drag calculations in the transonic regime, and of course I plan to continue work on the Vega model.
Work on the thrust curves has begun. I have introduced a new logical element for holding the curve data, so the curves will be displayed in the logical view just like regular vehicle components, and thrust curves are added to a model by dragging them from the browser onto an engine. So far it is possible to enter thrust value elements in the detail view of the curve, and the elements are displayed within the bounding rectangle of the curve feature. The feature also uses an interpolation algorithm to draw a smooth polyline through the elements, but the algorithm I currently use is unsuitable for this kind of curves. So I'll look for a different one.
ORT 1.3-0 is now avaliable for download. It comes with an up-to-date model file and a reworked modeller manual.
The model file now contains the first 3 stages of the Vega (P80, Zefiro 23 and Zefiro 9), along with the reference data for altitude and velocity taken from its debut launch. The simulation is completely off, most likely because the specified thrust values for the stages are maximum thrust values, not average thrust values. So until thrust curves for SRMs are implemented, the model will not be of much use. Next I'll bundle the current functionality into release 1.3 and then go after the thrust curves. I currently have no idea what the thrust curves of the Vega motors should look like, so I'm grateful for any pointers.
Yesterday I extended the Structure component with reference data sets, dealt with that in the model persistence and provided support for displaying those data on top of the logs generated by a launch. Reference data can be specified for altitude, velocity and acceleration. Because there is no editing support for the reference data as of yet, I manually carried the reference data table for the V-2 from Sutton's book over to the model XML. To my surprise, the actual simulation fits the reference data quite well. One issue I see is the matter of reference frames used by the reference data versus those used by the loggers, so more work will be required to log data in different reference frames. Next I'll add UI editing for reference data, then move on to additional loggers. I also itch to model the Vega, seeing that it had a successful debut launch (including the reference data provided by http://www.youtube.com/watch?v=j_Xk7wzlF4o).
I picked up ORT development again, this time on my Ubuntu box. This provided an opportunity to fix the font handling of the diagram view. Prompted by user feedback I also improved the error message when the modeller fails to load a model file. I'm currently lugging around Sutton's "Rocket Propulsion Elements", so the upcoming work will most likely be related to engine behavior (improved thrust calculation over atmospheric pressure ranges). Today I also did a very quick implementation of a polynomial interpolation routine (source: "Numerical Recipes in C++") which will find its use in thrust curve interpolation and probably a future version of a standard atmospheric model. Sutton's book contains additional details about the V-2 which I would like to include in the model file, however, this will require introduction of reference values overlaid on top of the simulated performance curves.
Work on the curve views is coming along nicely. The trunk now provides a menu for toggling side panes and the new south pane where the curves are located. A launch provokes an update of the curve views, which currently detail velocity, acceleration and altitude over time. The curve view automatically determines local maxima and minima and outputs rectangle features to visualize them. I did quite a bit of refactoring to move shared functionality down into the canvas panel, which is the common denominator of all graphical views. The next goal now is to re-use the curve views for modelling solid rocket propellant thrust curves and "Cd over mach number" curves. I'm also currently reading "Rocket Development" by Robert H. Goddard, which will hopefully help in refining the Goddard 1 - 3 models.
Release 1.2.0 is now available in the download section. It contains mainly usability improvements and bugfixed associated with the Logical View, which should now behave more civilized in most every way. The model file version was increased from 1 to 2 to provide migration of existing diagrams to a 10-fold increase in scale, which became necessary for technical reasons. Apart from the fixes there is now also a dedicated feature for attitude control jets, support for vertically oriented separation modules and, last but not least, tanks now show their load level as a shaded area within the tank. This release also contains the atmospheric drag improvement mentioned in the previous blog post. The modeller manual has undergone some work as well, and I will try to continuously improve it on a per-release basis.
Today I committed an improvement of the calculation of atmospheric drag.
Before, the atmosphere of the celestial body was treated as a stationary gas, regardless of the motion of the planet's surface caused by the planet's angular velocity. As a result a vehicle launched from the equator suffered a drag corresponding to a velocity of roughly 463 m/s right from the point of liftoff, even though the relative motion of the vehicle actually starts out at 0 m/s. ... read more
Today I worked exclusively on the modeller manual, which is included in the new Release 1.1. This release also contains some enhancements/fixes to the modeller UI and the new Engine Calculator, which is the first entry in the "Tools" menu.
I finished the first draft of the Modell B. The numbers match Oberth's predictions to some degree, but it's hard to tell whether the differences are due to Oberth's own simplifications, imperfections in ORT or due to some own assumptions I was forced to make.
One rather striking error in the book appears to be the initial acceleration provided by the Hilfsrakete, which is specified as 100m/s2 and imparts an excessive velocity on the entire stack. So I manually reduced the thrust to a degree that asserts the specified 500m/s velocity on ignition of the A.R. The focus now is back on ORT bugfixes and new features.
I quickly slapped together a build file for ORT and used that to produce the first release, which is now available on the download page. The models.xml in that "version 1.0" already contains early work on Oberth's Modell B. It cannot fly yet because the engine thrust forces are still missing.
Yesterday a copy of Hermann Oberth's "Die Rakete zu den Planetenräumen" arrived in my mailbox. I'm very much looking forward to including his proposed rockets in ORT and see how they perform.
Okay, after a few weeks of development I decided to put ORT on SF to mitigate the risk of losing all the work to a hard disk crash. If you want to know what ORT is currently capable of please check out the main page of the project wiki.