Newcomers to the investment casting market, specifically buyers and designers, face challenges in navigating a new territory of terminology, value elements, and market forces. The purchasing and design methods that have worked previously might suddenly have unexpected shortcomings. Also, it can be difficult to make decisions when your contact with the industry is limited, and a small error can result in a project being stalled or never realizing its full potential.

This paper aims to convey practical information about the procurement and design of investment castings by demonstrating quantitatively the consensus among industry veterans about these topics. A broad group of casting professionals has been empaneled to supply input, including:

•    Investment casting engineers and foundry managers 
•    Experienced designers and buyers of investment castings 

This is not a paper about negotiation or design of parts (though it has elements of those useful things). Rather, it is a practical paper about avoiding common pitfalls as a manufacturer brings a new casting to market.

Certifications 

It is critical that the foundry and the buyer have a common understanding of what is expected in terms of certifications. 

This should be clearly stated in the RFQ and PO. This topic is commonly covered in industry standard specifications, such as those used in aerospace applications. That said, depending on the industry, the designer/buyer might want to evaluate what is needed for the individual application.

Certifications can cover chemistry, hardness, tensile testing, microstructure, and other metrics. Some of this testing is done at the foundry and other testing might need outside lab work. 

Some certifications and lab work might need to be broken out as a separate item on your invoice. Whether the tests are done on-site or at an outside lab, there are real costs involved. This is another case where over-specifying can just mean extra costs.
Lastly, you should never assume any certification or test is being done unless it is specifically stated as being done. For example, do you really need a remelt chemistry to be included after melting a master heat? This would need to be specified.

Panel Comments on Certs:
•    One heat treat lot might include four separate melt lots. They are certified separately.
•    Traceability and date codes are a big deal between melt lots and finished product. This includes heat treatment.
•    If you have a certificate of conformance, that tells you it has met a standard. This is different than inspections.
•    If parts are serialized, then the certification and all traceability needs to tie back correctly.
•    Be very specific about the level of detail required in the certification report. For example, you might need an oven chart from a heat treater.
•    You do not want untraceable certs when problems happen.
•    Some foundries will do certain tests routinely and other tests at the customer’s request.
•    Tests should be itemized out on the quote. New buyers tend to not understand this unless it is itemized.

Dimensions

The investment casting process is capable of achieving certain dimensional tolerances. These are not machine tolerances. Even so, you might not always need machine tolerances. 

There are different ideas about what to expect. In general, the typical tolerances supported by the investment casting process are:
•    Linear: 1in. ± .010 in.; add .005 in. for each additional inch.

Table 1 shows what is considered industry standard from the Investment Casting Handbook. Your source might be different.
This should be discussed prior to determining final design and machining requirements. The source foundry should readily provide what they are comfortable with.

Panel Comments on Dimensions:

•    Most foundries will publish their particular tolerancing limits on the website. I have seen as low as .004 in. per inch per inch but normally .005 in. per inch per inch for linear and profile dimensions. Anything tighter would have to be agreed to with the foundry and customer, which is why new parts should always be discussed with the foundry at the design stage to ensure all features are practical and cost effective for the manufacturer and customer. Over-specifying tolerances that cannot be achieved with the process are the cause of much of the price escalation and delivery delays seen in the industry. Tight tolerance features should have machine stock on them and be achieved by machining.

•    Early input is critical. Insisting on overly tight tolerances or properties after final design can mean the ruin of the entire project. We have seen casting go back to barstock for no other reason, and it could have been avoided if the designer had accepted foundry input at the early stage.

•    We will commonly have a fixture for checking the critical dimensions.

•    Every part is its own animal. Their geometries are different.

•    The dimensioning scheme on the drawing is very important, tying back to real casting features wherever possible and keeping the origin point as close to the center of mass to help keep dimensional tolerances. 

•    Your datum structure must match your GD&T protocol. Having a machined drawing helps, even if the foundry isn’t doing any machining.

•    Your datum structure should be friendly to both CMM (Figure 8) and scanning technology.

Surface Finish

Surface finish is a common selling point for those in the investment casting industry as compared to other processes. According to the ICI, the investment casting process typically supports maintaining a surface roughness of 125 RMS max. Mark White explains, “Surface finish requirements are usually implied in discussions and loosely referenced as an RMS or Ra value. Little discussion occurs about the functional reason for the finish unless it is an airflow requirement in a turbine engine casting. There is usually no supplementary discussion or agreement between the casting purchaser and the casting manufacturer beyond this.” The end user simply expects what they consider to be a superior surface finish compared to other processes.

To avoid future setbacks, surface finish should be specified on the print and/or RFQ. This includes indicating on the drawing the critical surfaces involved. The specification should be tied to a coupon or standard, such as the C-9 surface comparator (Figure 9). White goes on to suggest that the designer “Utilize ANSI/ASME B46.1 to write the surface texture/finish drawing requirement/casting specification.”

The designer/buyer should also understand that over- or under-specifying surface finish has a cost. That is, demanding a smoother finish than the part actually requires in service means devoting resources to meeting this requirement, and there are costs to those resources. These costs might not be obviously quantifiable; as such, it is wise to consult the foundry as to what is attainable at what cost.

With that in mind, the buyer should also know that casting surface finish can vary depending on the source (between different investment foundries). If dual sourcing, this can lead to parts that are functionally identical but possess different appearances.

Panel Comments on Surface Finish:
•    It makes little sense to demand or pay extra for surface finish you don’t need. This adds to your cost and to your supplier’s cost. It is easy to spend $1,000 on a $20 part.
•    We once had to buy a gauge for our customer to measure surface finish on our parts.
•    Gating locations can and do affect dimensions and surface finish. Ideally the gate will be on a surface that is later machined or does not affect fit, form, and function. 
•    There is a big difference between a tooled surface finish and a surface finish made from additive manufacturing.
•    It is challenging to measure. A good supplier will bring this up.
•    There are multiple ways to get the surface you need besides the mold cavity, but there are costs.
•    Surface finish standards and visual acceptance criteria are becoming extremely challenging in aerospace. Specific criteria need to be flowed down.
•    Surface finish requirements, real or perceived, often preclude the use of additive manufacturing, especially on internal features. 
•    Bad casting quality will lead to bad surface finish even on machined parts (latent defects).
•    We use the C-9 gauge; many foundries use it.

X-Ray Testing (Radiography)

Those new to casting quality might be interested in assuring a porosity-free part. With this in mind, they might specify X-ray testing of cast or cast and machined components as part of their quality program. Figure 10 shows an X-ray of an investment casting.
Generally, a sample size is provided in the spec. If there is a failure in that sample, then 100% of that lot (that specific view) is to be X-rayed. 

Designation of the X-ray class is important in keeping the price reasonable. Having a high-class designation requires 100% x-ray, when in reality you might only need to sample for your application.

Appropriate designation of X-ray grade is critical to the piece price due to the cost of the film and/or labor.

Careful consideration needs to be taken when applying NDT grades on machined castings as compared to raw castings. For example, “machined casting grade B raw casting grade C” is a problem waiting to happen. In other words, why would you read the X-ray of a cast part to the standard of a machined part?

If you allow digital technology instead of film, this can be a path to reduced costs.

Panel Comments on X-ray: 
•    Outside lab techs have difficulty reading casting results.
•    You need to know what you need and apply it; there are standards to help you.
•    The thickness of your part matters. The foundry might be able to readily X-ray a part three inches thick. They might outsource a thicker part.
•    Specifying “high soundness” areas while maintaining flexibility elsewhere can be an avenue to savings. 
•    We typically do fluorescent penetrant inspection (FPI) on most castings as well. FPI will detect cracks, surface oxides, etc.
•    X-ray cannot find everything. Also, you should specify it correctly to get what you want.
•    The purpose of the X-ray should be detecting conditions that affect fit, form, and function.
•    Digital and film X-rays are not the same; do some research before switching.
•    The Grade you specify drives costs.
•    Insure that any requirement for external approval for X-ray technique is specified in the PO/quote package. This could affect lead time.
•    A machining drawing will typically include the acceptable grade for the machined part; it should match the cast part.
•    A machining drawing will also tell you which areas don’t matter (they might be getting removed).
•    Machine shops will typically only do NDT on the machined areas.
•    A partial sampling inspection plan (common with X-ray) might still lead to problems after machining. This is commonly called a latent defect.

First Article Samples

All parties want the initial samples to be produced quickly and at minimal cost. That said, they need to be meaningful. Your first article procedure should be reflective of your stable process with appropriate sample quantities representing regular variation. At the very least, a full set of mold cavities needs to be represented. 

Also, different industries have different practices and customs. In investment casting, it is normal that these metal castings are paid for, even if the part is being sectioned for quality/inspection purposes.

At this stage, there is common precedent for a determination of parts being acceptable, with some detail out of spec that all parties agree is meaningless (maybe the overall width is 0.005 in. too great on paper, but functionally irrelevant). Parts go into production with a side goal of changing the drawing to reflect the new reality. However, people can get busy, and the drawing is never updated. Eight years later an engineer discovers 14,000 useful parts are out of specification. The point here is that when a drawing is supposed to be updated, even on something seemingly unimportant, then it should be updated expeditiously. Also, until the drawing is updated, a deviation should be provided. 

This is also a common time for another specific mishap. If sample parts are being sent to a customer or subcon vendor, those samples should be shipped to a specific person at a specific address, with a tracking number. Many companies have multiple addresses, and samples have been known to get lost.

Lastly, surface finish of prototype parts made by rapid prototyping might not meet surface finish specifications. Chances are, there will be some dimensions that are noncompliant also. All parties should keep this in mind.

Panel Comments on First Article Samples 
•    A good practice is to use a box that is brightly labeled such that it draws attention.
•    Most buyers/designers think of first articles as “dimensional.” In reality, it is also an approval of metallurgical soundness.
•    “Not correcting the drawing” can be a very expensive experience for all involved, up to hundreds of thousands of dollars.
•    Many SLA patterns will affect the surface of the shell.
•    Updating drawings is required per ISO.
•    You should clearly differentiate between prototypes and first articles.
•    If you are going to use scanning technology for inspection, that needs to be agreed up front.
•    This step tends to hold up the entire process.
•    Samples get lost all the time. Did they go into production?

Hidden Costs 

There can be some costs that are incurred in the investment casting process that are not readily apparent. Some of these are readily avoidable.
Relocating a part. Sometimes, in the course of a business relationship there comes a time to move a component from one source to another. There could be any number of reasons. When doing this, the buyer should consider:
•    New tooling—is the existing tooling compatible with the new source’s equipment? Sometimes it can be reworked or adapted, sometimes it cannot.
•    Requalification—will your customer require engineering time and resources to requalify the part from the new source? The PPAP process, in particular, can be expensive.
•    Production or service status—is the part going from full production to service status necessitating the move? If so, then many factors will be changing, and this will affect pricing.
•    Engineering time—even without requalification, will you require internal resources to make the move happen?
•    New pricing upon return—if the move fails, it is possible that the old pricing will not be honored upon return to the original source.
•    Shipping—freight, brokerage, and time in transit are quantifiable costs that need to be considered.

Surcharges. The raw materials inputs used by foundries can vary in price considerably over time. These inputs include things like energy and alloy (nickel, molybdenum, etc.). 

Some foundries (especially ferrous foundries) structure their pricing to have a base price and a surcharge based on these raw material price fluctuations. Some do not (Figure 11). When comparing the price of castings between two foundries it is imperative that these surcharges are calculated correctly. If surcharges will apply, then the conditions under which they will apply should be agreed upon up front.

Surcharges are generally a non-factor in nonferrous foundries using certified ingot.

Subcon markups. Suppose the buyer is buying a component from a foundry, and the foundry is providing that component cast, machined, heat treated, and painted condition. The casting and machining are done in house, and the heat treatment and painting are subcontracted. 

That foundry will mark up the subcon operations (heat treat, painting, etc.) if the internal capability doesn’t exist. When you purchase a heat treated and/or machined casting, the typical arrangement is to let the foundry handle the subcontracted vendor. They normally have a network of reliable suppliers set up. 

The vendor will normally mark up the cost of these operations by some percentage. This is common and acceptable behavior; the foundry is managing the vendor, carrying the dollars, and taking responsibility for the quality of the part.

If the buyer wants to avoid these markups, the foundry will usually be glad to provide a raw casting to be processed by the buyer’s own network of service providers.

Quantity requirements. At the time of this publication, the investment casting industry looks to be busy for the foreseeable future. Steady flow is important, as ramping up with additional labor is not realistic in contemporary labor markets.

 With these things in mind, a distinction needs to be made between whether the buyer needs 10,000 pieces all at once or 10 pieces per month for 30 years. A one-time 10,000-piece surge will require a premium because you would be overwhelming the source. 

Just-In-Time. Also, Just-In-Time is an important concept. But in many foundries (and subcons), there are set-up fees. Paying short-run premiums at the foundry, heat treater, and machine shop can overwhelm the savings of making extra parts and carrying inventory.

PPAP. Lastly, the PPAP process is time-consuming, and many foundries (and subcons) charge a fee for these documents. This is sometimes, not always, broken out as a separate line item on the invoice/quote.

Panel Comments on Hidden Costs:
•    When relocating a part, you might see significant dimensional differences simply based on foundry wax and sand choice.
•    Stocking raw or finished castings has its place.
•    When tools travel from shop to shop, their condition might change.
•    Some portions of some tools are considered intellectual property of the foundry. Removal of these portions means mutilating the tool.
•    Just because something isn’t itemized doesn’t mean it’s free.
•    You might be better off just ordering a new tool. We all have different waxes, different shells, etc.
•    The standard of the industry is shipping +/- 10% on shipments unless agreed otherwise.
•    Economical order quantities should be negotiated up front with the foundry.
•    Modification charges are common when moving a tool. For example, shrink factors might be different.
Additive Manufacturing And Investment Casting
Additive manufacturing (AM) technology changes rapidly, with advancements being made on all fronts. There are many ways to use AM in investment castings:
•    Printed resin or wax patterns.
•    Printed tooling.
•    Printed fixturing.
•    Printed cores.
•    Printed molds.

In the context and timing of this paper, you can and should readily prototype with AM technology in any of the applications listed above. Figures 12 shows an example of a tool made by additive manufacturing, and Figure 13 shows a rendering of the wax pattern made from that die.

Some designs make sense in AM for full production (as in printed waxes). If you are doing a short run of a part with any complexity in the tooling design, you should seriously consider AM. 

Also, time is money. Tooling for wax patterns can be printed in a short time frame, as well as the patterns themselves. If you need parts in three weeks, you should probably just go straight to printed patterns.

That said, in other cases you are better off just building a tool. It is easy to think, “this is the future,” and in some cases it is, but certain designs just make more sense economically with regular tooling.

The choice gets down to straight arithmetic regarding:
•    How complex is the part?
•    How many do you need?
•    How soon do you need them?

Panel Comments on Additive Manufacturing:
•    You need to weigh AM case by case. It is not always obvious. As simple part conventional tooling might be cost competitive. On a complex part, AM may or may not deliver the surface finish or tolerances needed. On long running things that don’t change, conventional tooling has a high up-front cost and much lower later costs.
•    You need to know how your additive supplier prices the work. Is it by the dollar per cubic inch? Or machine time? 
•    You can’t do AM without a CAD model.
•    Having your own machines (like for SLA models) makes you more competitive.
What To Include In The 
Purchase Order
The purchase order should reflect the same items covered in the Request for Quote, such as:
•    Part Number
•    Drawing Number, including Revision Level
•    Quantity Ordered
•    Purchase Order Number
•    Piece Price
•    Date Required
•    Buyer’s Name
•    Ship to Address and Bill to Address
•    Material Certification Required
•    Heat Treat Condition Required
•    Any Other Certification or Testing Required

In addition, real-world experience has shown that packaging requirements can become critical at this point. Expensive parts can be damaged late in the manufacturing process.

For example, when parts are sent for heat treatment or machining, the subcon will often process the parts and repackage them in the same packaging in which they arrived. This may mean less than ideal protection for finished parts.

Panel Comments on the Purchase Order:
•    We commonly have special packaging for high volume jobs.
•    Special packaging considerations should really be covered at the RFQ.
•    The PO is the overriding document. It should include the quality requirements
•    Including the wrong revision of the drawing in the PO is a common mistake. 
Conclusions (from entire paper)
•    A clear consensus exists across the Investment Casting industry, including the customer base on some critical topics in component purchasing and design.
•    Investment castings can often offer an avenue to cost savings for OEMs as an alternative to fabrications and forgings. 
•    Customer support makes a huge difference in the successful conversion from fabrication/forging to casting.
•    The investment casting process offers multiple creative methods to optimize component design and material.
•    Tooling quotes need to be scrutinized correctly.
•    Open and early communication between the buyer and the potential supplier is critical to maximize savings and minimize costly missteps.
•    Cooperation can also help to reduce lead time.
•    Clearly specifying realistic quality requirements for fit, form, and function is recommended. This includes test bars, surface finish, and X-ray testing.
•    Dimensional capabilities vary between foundries and often from job to job, depending on geometry.
•    Drawings should be updated when they need to be updated.
•    There can be hidden costs in the investment casting process. They can often be avoided or minimized readily.
•    Additive Manufacturing is a powerful tool in the investment casting toolbox. It should be considered and applied wisely.
•    You might want to think about packaging for your finished parts.
•    A condensed version of this paper is available in Appendix 1 for easy reference. Some edits have been made for brevity and space concerns. In case of any conflict, the original verbiage supersedes that in the Appendix.