Category Archives: A 3D model of the Venetians ships (2014/MP1)

Progress Blog Post 3: refinement of details and first quantitative results

Since the last intermediate report, our ship grew until reaching its final shape, so that the assembly and modelling of the Galeazza’s parts is now complete. In this third intermediate report we will present the final 3D model of the Venetian galleass and the problems that occurred during the last modelling stage. Then, we are going to discuss about the advancement in the rendering procedure, as well as the progress we made in understanding the rendering software Maxwell Render. Afterwards, we will expose the first quantitative data that we were able to gain from the 3D model, including some estimations and calculations that might result very interesting under a DH perspective. Finally, we will focus on the future goals where we present further information and quantitative results that we plan to acquire until the final report and presentation.

Final modeling

The modelling of the ship is finally done and we are really glad about the impressive result. Since the last intermediate report we modified the sails and added a lower internal deck with the different materials that could be found inside. One of the main purposes of our project, in fact, was to model the internal structure of the ship, as well as some of the loads and tools that were likely to be found in the hold. This will permit us not only to estimate the load distribution and the capacity of the Galeazza, but also to have a taste of life on such a ship by rendering what a person on board would see. Therefore, after creating different volumes in the lower deck, we filled them with materials and tools according to the information that we have gathered from the literature [1]. Besides the internal part of the ship, also several details like anchors, fences and ladders have been added throughout the model to fulfil the appearance of our galley. Last but not least, we added the oars to finalise the interior of the hull.
In the following video it is possible to appreciate a 360° external view of the final 3D model.

The size of our 3D model file has now reached 376 MB. Since now our laptops have been able to deal with the computing power demanded to model the ship. However, when we started to render the different layers and model repetitive parts, we faced problems to work properly on our model because of the huge amount of memory consumed. Therefore, a more powerful working machine would be necessary to try out some other material or weight estimations, as well as to run simulations and render videos.

Rendering

In parallel with the modeling of the last parts of the ship, we went along with the rendering process. In order to produce a final result as realistic as possible, we had to split this procedure in three different stages:

  1. Select a specific piece or layer of the galleass and consult the literature to identify the material it was made of.
  2. Browse the Maxwell Render online database in order to find the correct material.
  3. Apply the texture to the selected piece and adjust the roughness, transmittance and reflectance properties until the rendered part looks photorealistic.

Up to now, we managed to render successfully the three main external decks of the ship, as well as many objects and pieces composing its internal structure. The following pictures show the rendering of some parts after one hour of processing.

rendering Cannon Render AnchorFigure 6: Renderings of some parts of the ship. Top panel: Upper external deck with wooden protections. Middle panel: Cannon positioned on the intermediate deck. Bottom panel: The anchor and the rope holding it.

From the images reported above it is possible to appreciate the great level of detail that we were able to achieve, especially in the reflectance and light scattering of the different surfaces. Despite the quality of the images is quite good, it is possible to produce even more realistic results by increasing the processing time of the Maxwell Render software. However, due to the amount of images that we plan to produce and the limited computing power that we have access to, we decided to leave the processing time to one hour per image.
Besides the application of textures to the parts of the ship, we also had to create an environment in which the model could be exposed. Therefore, we reproduced several sky domes surrounding the ship in order to render the full model under different lighting conditions (Figure 7, top panel). Moreover, we decided to replace the sea that we presented in the previous blog post with a more realistic model of the water obtained with the tool Maxwell Sea. This extension for Maxwell Render permits to simulate realistic ocean waves from a simple plane just by setting a few parameters (wind speed, wind direction, water depth, …). The sea that we were able to simulate in this way is shown in the bottom panel of Figure 7.

ski seaFigure 7: Example of environments created to present the model. Top panel: Sky during the sunset. Bottom panel: Ocean with a wind blowing at 30 m/s.

First quantitative results

Having a detailed CAD model of the Venetian Galeazza allows us not only to produce nice renderings of the exterior and the interior of the ship, but also to obtain quantitative results that might be appealing to naval experts and the Digital Humanities community. The first data that we have been able to estimate are reported in the following table and commented below.tableTable 1: First estimated data of the Venetian galleass.

The length, the width and the height shown above refer to the maximal quantities, meaning that they take into account the aftercastle, the frontcastle and the masts.
The total sail surface is 1280 m2: this splits up to 283 m2 for the Trinchetto, 358 m2 for the Mezzana and 638 m2 for the Maestra sail.
Knowing that those ships were usually built from walnut wood with a density of approximately 700 kg/m3, the total mass of the empty ship can be estimated. The volume of the used wood that we extracted from the Rhino model is equal to 580 m3 which yields a total mass of about 400 tons for all the wood of the empty ship.
To keep the ship upright, cannonballs were placed on the lowest empty deck of the galleass to serve as ballast (as we can observe from Figure 4, previous report). In order to derive the maximum number of cannonballs that could be stored in this space, we had to compute the average ball size. According to [3], a typical canon ball of the epoch was between 5 and 30 lb heavy: using a weight of 12 lb (5.4 kg), we could derive that the average ball diameter was 11.2 cm. Given this size, the CAD model allowed us to estimate that the lowest empty space could store up to 3 layers of 8000 cannonballs each, which sums up to a maximum of 24000 cannonballs, with a corresponding weight of 131 tons. As this is already about 33 % of the mass of the empty ship, we could conclude that this volume was never completely filled.

Future estimations and work

Besides the data reported above, we plan to investigate further questions and come out with other quantitative results. In particular, the points we will explore in the next days are the following:

  • Position of the waterline: this information is useful to know which portion of the ship was underwater and, hence, protected from cannonballs.
  • Centre of gravity.
  • Volume of stored water: after completing the modeling of the internal part of the galleass, we noticed the presence of a narrow empty space running under the rowing deck and along the whole length of the ship. After examining the literature [2], we discovered that this space was used to store potable water in boxes. We plan to precisely compute the number of liters of drinkable water that this ship could carry.
  •  How long the ship could stay away from harbor based on water and food storage.
  • Volume that can be used for merchandise when the galleass was no longer used for warring purposes.

We will try to assess the first two points by using the Orca3D simulation software, as explained in the previous blog post. Moreover, by the final deadline we plan to produce a series of rendered images both of the interior and the exterior of the ship under different lighting conditions. If we are able to access more computational power by the end of the project, we will also render a video of the galleass for a better visual impact.

Milestones (updated)

cds


References

[1] Guido Ercole, Vascelli e fregate della Serenissima. Navi di linea della Marina veneziana, Gruppo Modellistico Trentino, 2011.
[2] Guido Ercole, Galeazze. Un sogno veneziano, Gruppo Modellistico Trentino, 2010.
[3] http://netwar.wordpress.com/2007/08/15/the-battle-of-lepanto/