Economical Manufacturing Technique

IMHO - due to the limited quantity, you should try to find a suitable housing that is already manufactured by another company. A company already producing this part be it sand cast or lost wax process would have the equipment and experience to produce a quality product which will save you set up and start up costs.

As far as a choice between sand or lost wax casting, that would depend on the application.

Given my experience in the investment/lost wax process, I would think that sand casting would be the less expensive manufacturing means as you can remove the cost of a centrifuge from the process.

The lost wax process will give you a harder, more rigid product over sand casting.

As far as the turbine blades, yes, lost wax is the way to go.

To add to what has already been said, there is no economical way to cast 30 parts of anything. That's far too low a quantity. And you have two parts of low quantity. Even worse. So perhaps there is a compromise you could consider. Purchase one of the parts... whichever one you can find closest to your design as possible, and design and cast the other part to fit the purchased part. Your savings will be considerable.

As an alternative, you might look into the cost and appropriateness of stamping or pressing the housing out of sheet metal, if you simply must have a particular custom design.

If you want good information from an actual engineering background in the manufacturing of centrifugal pumps you will need to provide some additional information. I'm glad to help you on this; but I do not have time to write a generalized text for you.

1. What is the what is the head and capacity as well as liquid type, temperature and inlet pressure to be pumped if you have not already determined the materials of pump construction?

2. What speeds are available from whatever motor/engine you plan to use? From this information and #1 I can determine specific speed and general size of the pump parts.

3. Where do you plan the manufacturing of 30 units to take place? Like which nation? Capabilities vary from nation to nation and I don't want to suggest a method not available to you. This is very important. If you feel a need for secrecy email me your location individually and I will respect your situation.

I can't imagine you using any forgings unless you chose to manufacture the pumps entirely out of steel and welding pieces together. Such an approach might make some sense for a quantity of less than 5, especially if skilled welders were available at low wages, cost of materials were a large consideration and corrosion was no consideration.

Ed Weldon

Ed, from what he wrote, I was under the impression that he has the sizing, speeds, parts, capacity, and general design, under control; but only needs advice on an economical method to manufacture the housing and turbine in those low quantities.

But, I've been wrong before...

OBE -- The size of the pump has a lot to do with how you are going to make the pieces. The OP has given no clues on that. The pumped fluid and it's temperature and viscosity determine minimum allowable materials. Materials are also determined by manufacturing techniques available and size effects most economical materials. Specific speed determines the general geometry of the pump which has a second order effect on materials selection. Open impellers, easier to manufacture but less efficient than closed impellers are needed when insoluble solids may be present in the liquid stream.

What I forgot to ask was a description of the design level of the pump at this point and what engineering and design capabilities the OP has at his disposal. A full set of engineering drawings is more useful than no design information at all. If an existing pump is to be copied exact, that offers its own constraints. If engineering resources exist to create solid models of the design and the available manufacturing technology is capable of importing them the door is open to less expensive short run manufacturing techniques both in the foundry and the machine shop.

This is indeed a complex subject that a whole book could be written on. To my knowledge none such has been published in my language, English. And I do not have the bandwidth to do such a project or wide range of knowledge to cover it in a comprehensive manner.

Generally a centrifugal pump for water service in the range of 5 to 500 HP, flow conditions from 100 liters/minute to 10,000 liters per minute and total developed head of 10 to 200 meters are traditionally tooled and constructed in the following manner for quantities under 100 total:

1. Casing from two pieces, sand cast iron class 35 gray iron from wood patterns on match plates and wood core boxes. Green sand for outer mold, baked oil sand cores. Machining of casings is typically done on an appropriate sized engine lathe or turret lathe. Holes can be done by any drilling and tapping equipment capable of accurate hole location. Post machining pressure testing by water at pressures 150% of maximum working pressure are appropriate. Un-machined inner wetted surfaces of the casing are typically painted with a water resistant coating to deter rust when the pump is not operating and air is able to remain in the casing for prolonged periods.

2. Impeller from cast iron as above for open impellers. Closed impellers baked oil sand cast bronze from wood core boxes with careful attention to post casting cleanup of fins and other casting anomalies. Typically for fresh water service bronze alloy is so called 85-5-5-5, CDA 833/836. Impellers are typically machined to final dimensions on an engine or turret lathe.

3. Shafts machined from bar. AISI 416 stainless steel on smaller pumps with no shaft sleeve machined from bar. AISI 1040 not hardened steel from bar on larger pumps with sleeves. In all cases close tolerances on shaft and housing diameters at ball bearing mountings should be as recommended by ball bearing manufacturers. This requires a high skill level in the machinist and adequate measuring tools especially in the larger pumps. Grinding on centers is commonly employed for finishing shaft bearing mount surfaces. A milling machine operation is typically required for shaft keyways on larger pumps.

4. Additional wetted shaft fittings such as sleeves, impeller nuts, washers, stuffing box covers packing lantern rings and impeller keys are usually either stainless steel 416 or 304 or one of the higher strength bronzes in the CDA 400 or 500 series like CDA 465 or 485. These are typically lathe machined from bar stock.

5. Non wetted parts like bearing frames can be any combination of carbon steel AISI 1018 or 1020 and cast iron class 25 or better or aluminum for non rotating parts.

5. Bearings are typically grease lubricated ball type isolated from water splash by rubber lip seals. Ball bearings typically lubricated with the same intervals and methods as electric motor bearings of similar size and duty. Milling machine and lathe operations are typically employed.

6. Mechanical shaft seals are preferred for ambient temperature water service. Specific application information should be supplied to the supplier of the seal. Mechanical packing can be effectively used on most pump applications but require frequent attention by trained maintenance personnel to insure acceptable cooling flow levels.

Maybe this will help. But I really need information from the OP to make specific recommendations. Generally investment casting is not employed in manufacturing pumps in small quantities. However there may be reasonable exceptions for smaller pump parts where existing parts are being copied and a good machinist or pattern maker can economically build up certain surfaces to provide casting shrinkage and machining allowances. Since wax molds are relatively inexpensive for smaller parts and are generally made from a male master that looks like the final part this method could make sense. The only materials that generally don't lend themselves to investment casting are the cast irons and low carbon steels. In the USA CF- and 17-4PH stainless steels, CDA 836 brass are most common investment casting alloys. Ditto aluminum 356 which may be a possibility for small pumps under 5 hp.

I suppose you could build a pump by welding up premachined and formed pieces of steel and silver brazing brass for the impellers. Very labor intensive and would require fixture building labor almost as much as making wood patterns. I can't see many places in the world where this would be practical. Usually where you find lathes you will find a foundry of some kind.

Ed Weldon

Well... you've certainly got a handle on this.

Granted, what he's trying to do is fairly odd. But, I guess it's a big world out there.

We don't know the size of these parts but if not too big an option for the casing might be to have a 3D printer produce wax patterns for investment casting.

That's if a solid model is available. For all we know this guy may have nothing more than an idea or an actual pump to reverse engineer. BTW a solid model can also be used to save a lot of time in shaping a wood foundry pattern, foam molds or actual foam for lost foam castings. Anywhere that physically large high tech consumer products are being manufactured and packaged for transport there is going to be manufacturers who work with low density foam molding. "Lost foam" is quite adaptable to cast iron foundry work.

Ed Weldon

ED thank you so much for your valuable input and sharing your experience , i have taken note of all what u said and would request more of your assistance after finalizing few details

regards

Agreed, foam patterns work well for small cast runs.