Barriers to exploitation – the need for Innovation
There are key technical barriers which prevent the immediate exploitation of the proposed HeatSculptor design and production route, these are:
- Modelling of fluid flow and heat transfer: Only recently have powerful modelling tools become commercially available that enable both fluid flow and heat transfer to be computed; exploiting this capability in the design of heat exchangers is in its infancy.
- Heat Exchanger Design Limitations: The design of heat exchange surfaces has been limited by the capability of available established surface manufacturing processes.
In principal, the innovations to be developed within HeatSculptor that will overcome the above barriers fall into three areas:
- Modelling of fluid flow and heat transfer
- Manufacture of complex surfaces for heat exchange
- Transfer to, and validation of, the ‘Surfi-Sculpt®’ technology in a production environment
These innovations and the technical barriers required to be overcome to deliver HeatSculptor are summarised below, along with the partner who will be delivering that innovation.
|Heat exchanger designs are limited by currently available surface manufacturing processes|
|The new EB surface structuring process allows a greater freedom of design for surfaces||An understanding of heat exchange/fluid flow and hence the knowledge of optimum heat transfer features||CEN|
|The manufacture of heat exchangers using the Surfi‐Sculpt process has not been fully developed and quantified|
|Novel surfaces for efficient and improved heat exchange||Surfi-Sculpt processing experience||TWI|
|Heat exchangers have historically been designed empirically and effective modelling is in its infancy|
|Modelling to optimise the new range of surfaces available||Access to, and experience in, use of CFD to model complex fluid flow and heat transfer problems||CEN|
|Surfi‐Sculpt technology has not yet been exploited in a production environment|
|CAD/CAM and HMI Software for improved understanding and simple operator use||Electron beam interface and CAD/CAM software knowledge||TWI|
|Process is not currently cost effective to deploy in a production environment as only a small area can be worked at a time|
|Wide angle deflection allowing larger processing area||Expertise in electron beam equipment and processing||TWI|
Modelling of fluid flow and heat transfer
Specific modelling packages are required to solve complex heat transfer problems and interpret the results for assessing the surfaces of products against quality control guidelines. Current practise is to design heat exchangers empirically and within the capability bounds of available manufacturing processes. The new Surfi‐ Sculpt processing route enables greater emphasis to be placed on the heat exchanger design process.
In order to assess designs and generate optimum surfaces, it is most efficient to adopt a modelling approach. Due to the transitional nature of the flow and the complex geometry which is foreseen, large scale turbulent problems are generated. This type of flow requires the use scale-resolving modelling strategies, ie direct numerical simulation (DNS) or large eddy simulation (LES) need to be used, instead of the standard averaged RANS (Reynolds averaged Navier-Stokes) turbulence models in order to obtain accurate results. Due to the high accuracy and resolution requirements of these strategies, it is foreseeable that commercial tools will not be able to provide the necessary resolution for accurate flow predictions. It is therefore proposed to use a research code based on novel unstructured high-order accurate discretisation’s (in this case the discontinuous Galerkin method). This method has been validated for DNS and LES for transitional flows, thereby outperforming classical methods, and provides furthermore a direct feedback on the resolution of the mesh. The latter property allows target mesh refinement and avoids heavy grid convergence studies. Due to the fact that these are relatively new advances in Computational Fluid Dynamics (CFD), the knowledge and experience in using them to solve complex problems is specialised and requires expert knowledge.
Use of such software will significantly enhance the knowledge of heat transfer design as it is the only way to calculate heat transfer over complex geometries. Used as stand-alone tools the user can test the heat transfer capabilities of a specific surface design, but in order to improve it, has to rely on educated guesses that need even more experience on the specific subject.
Therefore the modelling technique will be combined with the in-house generic optimisation suite Minamo, developed at CENAERO. This suite is currently used on a daily basis, mainly but not exclusively for the multidisciplinary and multi-objective optimisation of turbomachinery blading. By coupling the new flexible manufacturing process (Surfi-Sculpt) with this new advanced modelling capability, HeatSculptor will enable the design, optimisation and manufacture of efficient, high performance heat exchangers.
Manufacture of complex surfaces for heat exchange
In order to obtain maximum benefit from both the modelling and the surface engineering, integration of the two aspects is required. The Surfi‐Sculpt process is dependent on material properties and currently surfaces cannot be manufactured by the electron beam system directly from CAD drawings without operator intervention to achieve the desired surface features. The Heat‐Sculptor project will move towards this objective by generating an easy‐to‐use human‐machine interface and the development of CAD/CAM.
Surfi-Sculpt is a novel power beam materials processing technology which creates carefully designed features across a surface. A power beam is used in Surfi-Sculpt processing. This can be a laser beam or an electron beam (EB), although the latter is significantly more advanced in its development and has been used in the vast majority of tests to date. The beam is deflected rapidly over a substrate surface to displace material in a controlled manner. The result is a textured surface consisting of an array of up standing protrusions above the original surface and a corresponding array of intrusions or cavities in the substrate. The size, shape and distribution of the features can all be varied to produce a surface tailored for specific applications and heat transfer characteristics.
When the beam comes into contact with the substrate surface it creates a molten pool of material. If the beam is then translated sideways, the combined effects of vapour pressure and surface tension allow the material from the hole to be piled up behind the beam as shown below. By repeating this process many times at the same site, protrusions may be grown, and each will be accompanied by one or more corresponding intrusions or holes. A whole series of protrusions can be built up simultaneously across a substrate.
With careful control of the electron beam process parameters (beam accelerating potential, beam current, focus, etc), the design of a unique pattern and precisely defined deflection movements, it is possible to create a wide variety of different surfaces which can be tailored to different applications suitable for both batch or mass production environments, a few examples of which are shown below.
Examples of surfaces produced by laboratory Surfi-Sculpt processing: array of curvilinear spikes, honeycomb lattice, tube processed using a ‘travelling’ pattern of spikes to increase the surface area by over 50%
Surfi-Sculpt process in action. (courtesy of TWI – all rights reserved.)
Transfer to, and validation of, the ‘Surfi-Sculpt’ technology in a production environment
The electron beam system has limitations in the size of area that it can process using beam deflection alone and is typically applied for one‐off customised or small quantity production. In order to meet the needs required by high volume manufacture, Heat‐Sculptor will develop a wide angle deflection system allowing for processing larger areas and develop CAD‐CAM software in order to enable effective and efficient design-to‐metal programming for improved understanding and simple operator use.
The electron beam process is not currently cost effective to deploy in a production environment as only a small area can be worked at a time. Therefore, the production rate achievable by the electron beam equipment is inadequate and needs to be expanded in order to fully exploit and transfer the ‘Surfi-Sculpt’ technology in a volume production environment.
Electron Beam equipment is typically applied for one-off or small quantity production due to the deflection systems causing beam aberration that compromises the processing quality for processing areas any larger than 750mm2. Although, in comparison with machining, EB equipment can be economically viable, as non‐recurring engineering costs (e.g. special tooling for machined components) would be amortised. However, a wide angle deflection system allowing for a larger area to be processed needs to be developed in order to achieve the higher production rates required by high volume manufacturing and at the same time allowing the system to remain cost competitive.
The translation of design-to-metal is currently challenging for the electron beam process and requires highly skilled and knowledgeable machine operators. The development of CAD/CAM and HMI software is required in order to enable effective and efficient design-to-metal programming and improved understanding to achieve simple operator use. This development would include the improved translation via semi-automatic machine programming; especially where curvilinear surfaces must be processed with wrap-around features. To achieve this, an improved understanding of how the beams interact with the substrate surface, and how the resulting molten metal moves, is required.