Building Prototypes of the Kreios G1 Gobo Image Projector for Osram.
In 2012 Cambridge Industrial Design approached Solve 3d to undertake a special project for their client Osram. We were tasked to produce a set of 4 prototype projector lights, to reference level that would perform safely & reliably at a lighting exhibition. The prototypes were to allow Osram to confidently demonstrate the unique features of the product.
Running continuously for 8-10 hours per day, the high output projector generates significant heat. The heat is dissipated through the aluminium body where the increased surface area provided by the vanes allows greater airflow. The airflow of course, acts to cool the surface and maintain a working temperature, essential when plastic parts are in close proximity. In a prototype, safety and reliability has to be guaranteed.
In manufacture, the choice of materials and processes is dictated by the function & aesthetic of the product. At prototype stage however, there is no pressure die cast, extrusion or injection mould tooling. Prototypes of this nature are not just ‘Look/See’, they have to perform as the final product. Not only are they to test performance, they have be an accurate reference for the final product. These prototypes served as a powerful sales tool enabling Osram to take advanced orders, with customers believing that the product was already available.
It starts with CNC machining.
The aluminium end caps of the G1 were pressure die cast in production. For the prototypes they were machined and the extra deep and narrow vanes posed challenges that called for an unconventional approach. It is possible these days to 3d metal sinter these however the amount of finishing and the cost would be prohibitive.
If we were to machine conventionally, using long reach 2 & 3mm cutters, the process would tend to create rogue resonances that could lead to cutter and material deflection. This tends to result in gouges and an inaccurate part. The material resonances would ordinarily get worse as the space between the adjoining heat sink becomes unsupported. We solved this by pocket machining alternate ribs. By machining only one side of each rib, we left supporting material & regulating cutter feeds and speeds we minimised the resonance. Final finish cuts using a waterline path left perfectly machined sides with no gouges or vibration marks.
We then coated the machined surfaces with wax & poured fast cast polyurethane resin, filling the voids. The resin acted as a support for the ribs during the next stage of machining. Using plenty of coolant during machining the resin stays firm and supports the ribs perfectly. A perfectly smooth sided 3mm rib with 2.8mm gaps to a depth of 35mm was achieved. The next stage was to refill the machined side with resin, face cut the whole billet and then flip the billet over , locating it on 4 accurate ground indexing pins. The rest was reasonably straight forward machining’s.
The main body was a simple form and created using a more traditional machining process. This was the easier of the two main body parts and represented what became an extrusion in production.
The Plastic Components.
The plastic components needed to be heat resistant and self coloured. A paint application was too risky for the show so the decision was made to create master models with the requisite production finish and then to create silicone tooling for vacuum casting. Finishing an SLA was considered but in the end we decided to machine the masters from acrylic to ensure concentricity, and accuracy. In the past we had found that round SLA parts built on their side or at an angle ended up slightly oval rather than round due to cumulative discrepancies in the build layers.
The machined acrylic parts were more robust and required less finishing than would an SLA consequently controlling the cutting strategies, the total clean up time of the masters was around one hour. In the program we created a paint allowance of -0.15 mm. When the final VDI finish was applied the parts were acurate to within 0.05mm.
Following the application of the final finishes, silicone tools were produced and the final components were vacuum cast in self coloured polyurethane. A heavy mineral fill was added to give the castings greater resistance to the latent heat at the outer limits of the product.
Once the aluminium components had recieved an even bead blasted finish, a coat of high temperature etch primer was applied. This was followed by a high temperature top coat, that was post cured to creat a heat resistant shell. The more traditional route of powder coating was deemed to be too heavy duty for these prototypes.
After running quick test to ensure that everything was working correctly then all four projector lights were mounted to a test bed & were left to run for 10 hours. The units ran quite hot but when the temperature stabilised there were no issues so we were satisfied that the units were safe and ready to ship.
The Big Show.
Great Design & Engineering by CID combined with well-managed multidisciplinary prototyping, resulted in the projector light prototypes running flawlessly & safely for the entire exhibition. Clear communication between Osram, Cambridge Industrial Design and Solve 3D resulted in a very successful design & development project completed in less than six weeks.
This Red Dot award-winning product continues to sell successfully sold all over the world.