Minimal manifolds revisited - shedding more material & boosting performance

This post updates my previous Minimal manifolds post, reflecting a second design iteration by my colleague Kevin Brigden that sheds yet more material and further boosts operational performance by enabling a material change.  Firstly, a recap:

Design iteration #1

Manifold blocks are complex components where numerous pipes come together and intersect.  Traditionally, manifold blocks are created by cross drilling from a solid metal block.  Fluid passages tend to meet abruptly at right-angles affecting flow efficiency.  Additional components such as plugs are required, and the flow path design is compromised due to constraints of the manufacturing technique. 

The image below shows a fluid channel passing through a 90 degree bend, produced using cross drilling and end plugs. 

Computational fluid dynamics (CFD) analysis in the right hand image above shows areas of low flow (blue) which can result in a build-up of deposits, as well as some turbulent flow around the sharp corner (red).  One way of preventing clogging is to use further internal plugs (below), but these add complexity whilst not addressing the issue of the fluid losing pressure head by having to pass around a sharp corner.

So, from a fluid dynamics perspective, conventional manifold designs have room for improvement, plus they can also be rather heavy.

Additive manufacturing (AM), where we build up the part layer-by-layer rather than by removing material from a solid piece of metal, allows us to design optimised internal flow paths whilst eliminating unnecessary component weight. 

The example part is a substantial hydraulic manifold, currently made from a 25kg solid block of aluminium.  Let's look at how we can apply Design for AM (DfAM) principles to this part.

DfAM step #1 - extract the component essentials

The first step is to extract flow paths of the cross-drilled design as a starting point for optimisation for an AM build process.  Rather than starting with a solid block of metal, we de-construct the part into just the essential tubes that provide the functionality of the manifold block.  At this point, we remove any drilled areas that are not required for the flow path, leaving us with the pipe network shown in the right hand image below:

DfAM step #2 - optimise flow

Now we optimise the fluid flow paths, without the design constraints of cross-drilling, replacing sharp corners with rounded bends where we can.  The left hand image below shows a conceptual flow path, which is now ready for CFD analysis.  The right hand image below shows CFD results (generated using Solidworks Flow Simulation), identifying areas of flow separation and stagnation.  The flow paths are refined until we are satisfied.

DfAM step #3 - determine wall thicknesses and add supports

Once the flow paths have been optimised, wall thicknesses are generated based on either the customer's specification, or by using finite element analysis (FEA) stress modelling based on pressure readings taken during CFD analysis.  Refer to the left hand image below.

Finally, support structures are generated in the intended direction of building.  As shown in the right hand image below, support structures can act as both a scaffold to hold the component together, as well as supports and anchors for the AM build process.

The first design iteration yielded a part with a 52% reduction in part volume and a 60% increase in flow efficiency:

Design iteration #2

Design iteration #1 was built and evaluated.  The manifolds are used in series and need to be removed for servicing, and so some of the features in the original design were included to aid extraction of the heavy (25kg) block.  The AM manifold design enabled removal without the use of tools, and so an opportunity was identified to eliminate these extraction features.  Experience from the first design iteration showed a need to increase the part stiffness to avoid tool chatter during finish machining, so we considered a change of material - from aluminium to stainless steel - to facilitate this.

This second iteration resulted in a 79% reduction in part volume and 37% weight reduction, despite the switch from aluminium to steel.  Shedding material translates directly into a part cost reduction, as both the cost of materials and the time required to melt the metal is driven by the part volume.

The 60% improvement in flow efficiency was maintained, as was compatibility with the existing design.  The stronger material (316L) and single piece construction create a more rugged part with fewer defect opportunities. 

Possible further steps

Although not deployed in this instance, further efficiencies can sometimes be gained at this point by consolidating pipe fittings into the AM manifold design to reduce part count and assembly costs.

One constraint on this project is that the manifold should be a fit, form and function replacement for the original design.  This means that the entry and exit points for the flow channels are fixed.  With a clean sheet of paper, it is often possible to further reduce mass, increase stiffness and enhance flow by re-orientating and re-aligning the flow paths.

Summary

Manifolds offer great opportunities to apply sophisticated DfAM practices to produce high performing parts.

AM enables us to redesign manifolds to minimise material volume and optimise flow paths.  AM's flexibility means that we can use design tools such as FEA and CFD to the full to direct the design process. 

With careful design we can eliminate fixtures and minimise removable support structures from our design.  And we can eliminates block extraction passages, reducing part volume and space consumption.