Optimizing Press Operations

What is the ideal number of strands for each extrusion die?  In case of profile designs where it is possible to use one, two, three, or more holes in the die, what is the best number to choose?

A surprising number of extruders seem to decide as follows:  The first priority is to extrude the longest strands possible, according to the available cooling table length and the number of finished cuts that can be made from that length, of course allowing for stretching scrap at each end.  Then, if the billet size will permit additional strands, a multi-hole die will be considered; but in any case the number of strands will be a lower priority than maintaining the maximum strand length.  In this system, if a die could be made to produce either:

·       1 strand of 6 finish cuts

·       2 strands of 3 cuts

·       3 strands of 2 cuts, or

·       6 strands of 1 cut each

the decision would be for the longest possible strand --- one strand of 6 cuts.

While this system will result in the highest yield, or lowest percentage of scrap, it seems not to consider the total picture.  After all, scrap is only one part of the concern, and there are other priorities to be considered.  For the most efficient operation, overall, it is necessary to also take into account production rate, as well as the relative values of extra yield and extra production, to determine the best operation overall.  For this purpose the mathematical relationship of all the press variables has been examined.

Below are derived certain key equations for calculating the parameters needed for production optimization.  A spreadsheet model has also been constructed using similar calculations (Excel version 7.0); when the user inserts the key input data for his plant (into the shaded boxes) one may examine any set of assumptions one may choose.  The sample spreadsheet uses English units, but a Metric version is also available.

One example of the spreadsheet is also included here, with assumptions entered for all appropriate variables.  This example and many others which are representative of actual conditions have been tested, and the calculations clearly indicate the benefits from increasing the number of strands extruded per die, even if yield suffers.

FORMULAS FOR OPTIMIZING EXTRUSION PRODUCTION

DEFINITIONS

n            number of strands

x            number of finished cuts (pieces) per strand

l            length of finished pieces

s            scrap allowance each end of strand

V            extrusion speed

w            weight per unit of profile length

L            maximum stretch length (table)

b            usable billet weight (net of butt)

d            dead-time cycle

t            extrusion time per billet

T            total time per billet (t + d)

B            billets per hour

Derived Formulae:

Strand length = lx + 2 s < L   (strand length must be less than total table length)

Extruded weight = n (l x + 2s) w < b  (total weight extruded must not exceed                                                                                     the maximum usable billet weight)

Extrusion time per billet = t = (l x + 2 s)/v

Total cycle time per billet = T = t + d = (l x + 2s)/v + d

Billets per hour = B = (60 min/hr)(60 sec/min)/T = 3600/[(l x + 2s)/v + d]

Gross production per billet = Pg  = n (l x + 2s) w

Net production per billet =  Pn = n l x w

Yield = n l x w/(n l x w + 2 s n w) = l x/(l x + 2s)

                                                                        3600

Gross production per hour  = (B)( Pg) = ---------------------------  (n)(w)(l x + 2s)

                                                                        (l x +2s)

                                                                        --------------  + d

                                                                              V

which may be simplified as:

                                                                  3600 V n w (l x + 2 s)

                                                            =   -----------------------------------------

                                                                      l x + 2 s + d V

                                                            3600

Net production per hour                        =  -----------------------------------  (n)(l)(x)(w)

                                                            (l x +2 s)

                                                             --------------  + d

                                                                 V

which may be simplified as:

                                                            3600 V n w l x

                                                =  -----------------------------------

                                                              l x + 2 s + d V

--------------------------------------------------------------------------------------------------------------------------

Conclusions:  The following principles may be learned from these calculations:

Possible steps to maximize yield:

      Increase the finished cut length (however, this is determined by the customer)

      Increase the number of finished cuts per strand (as will be seen below, this may increase yield but will not result in maximum production or the most efficient overall operation.)

      Decrease the stretcher scrap allowance (this requires redesigned stretchers and efficient use of the puller)

Possible steps to maximize (good) production:

·       Increase extrusion velocity

·       Increase the number of strands

·       Increase the weight per unit of length (this is usually not an option, as it is determined by the profile design)

·       Increase the finished cut length (this is usually not an option, as it is determined by the customer)

·       Increase the strand length so as to increase the number of finished cuts per strand (mathematically, this is not as important as increasing the number of strands)

·       Decrease the stretcher scrap allowance (this requires redesigned stretchers and efficient use of the puller)

·       Decrease dead-cycle time (by modifying the press controls and hydraulic system)

·       Insure that press production is not “bottlenecked” by downstream equipment (usually the cold saw or stretcher)

·       Increase the useable billet weight (length) by installing a moveable ram on the press.

Most important is that in a large percentage of cases the extruder may profit by increasing the number of strands per die.  This may complicate the die design, and the die maker may prefer to take the easy way out; if he has his way the result may be higher yield but lower production and less profit.

Comments on the results in the example:

1.    The example assumes a gross profit margin (selling price minus billet cost =$0.45/pound) as the value of extra production which would be realized.  The cost to convert extrusion scrap back to billets is assumed to be $0.15/pound.

2.    Even though yields are extremely low as the number of strands increases, the dramatic increase in net production rate and profit margin will more than make up for it.

3.    These calculations do not include the effects of other operating costs on the overall profitability, but these will only improve with this new scheme.  True fixed costs will be spread over a much higher production rate.  Labor and supervision costs should remain fixed and will also be spread over a larger production base.  Variable costs such as water, power, repairs, and supplies should increase in proportion to production.  Overall, there is a good "leverage" to profitability --- profits will increase in far greater proportion to costs, and competitive position improves as well.