Wednesday, 31 May 2006

Automatic Assembly (7/16)

Any detailed investigation into the design of a product results in a more productive method of assembly. The parts and assembly operations used to put together the parts can be viewed with either manual or automatic assembly in mind.  When designing for automatic assembly, remember that the intricate feedback loop that co-ordinates human motors is not present in economically justifiable automatic assembly systems. For example, parts that are manually picked up the wrong way round can have their orientation corrected. The human assembly worker detects this error through sight or touch and quickly corrects the orientation.    Similarly, defective parts can be detected and discarded by a human assembly worker.

Automatic workheads do not detect rejects without the aid of complex sensor systems.  A defective part arriving at the workhead causes a jam and the workstation is down for a period of time. These events are minimized by having high component quality levels and re­structuring the inspection routines. Most manual assembly systems have three inspection stages - goods inward, during assembly, and upon final assembly of the product.  Parts or assemblies which do not fall within quality bands, at each of these stages, are rejected. Automatic assembly equipment requires higher quality components and, therefore, greater quality control is required at the goods inward stage than for manual assembly. Automatic assembly equipment, fed with high quality parts, gives a higher quality finished product than manual assembly.  The consistency of an automatic system, aided by high quality parts creates a high quality product.

Tuesday, 30 May 2006

Automatic Assembly (8/16)

Dedicated assembly systems, compared with flexible assembly systems, are generally intolerable of defective parts.  It has proved effective to install a 100% inspection station prior to the workhead to maintain up-time in some installations.

The ease with which a part can be presented to a workhead automatically at the required feed rate, and assembled within the specified cycle time, depends upon the; design of the part, design of the equipment, and the method of assembly. Each element of assembly automation has it's own maximum performance characteristic. For example, the cycle time of a pick and place unit depends upon the degrees of freedom, the actuator stroke and the actuating medium, i.e. compressed air, hydraulic oil, DC servo motor. A feeder  is limited by its maximum conveying velocity and the insertion process is affected by the positional accuracy of the rotary table, robot, or platen location.

Monday, 29 May 2006

Automatic Assembly (9/16)

The product designer must analyse the product and consider if it can be economically assembled in it's present state, or decide if design changes are necessary for automatic assembly.  In practice, the majority of products assembled manually require design changes to make automatic assembly viable.  The product designer must consider what further benefits can be gained from more re­designing of design features, assembly operations, or even the elimination of parts.  The process of design for automatic assembly is best effected by a systematic approach. A structured method for evaluating designs to identify inefficient features has been developed by UMass and other organisations.

An automatic feeding device and, at least, one workhead is required for each component to be automatically assembled into a part-built product.  A significant reduction in cost is achieved by eliminating a part from a product.  Designers should strive for the irreducible number of separate components per assembly, consistent with its performance and fitness for purpose.  An investigation into the function of the product exposes redundant parts and these should be eliminated.

Fasteners, which are separate from the component being secured, should be avoided.  Fastening technologies of the future are based on adhesives, ultrasonic welding, soldering, resistance welding, clip fastening, and twisted tab joining.  Fasteners can be classed as being permanent or semi-permanent. Permanent fasteners do not permit removal, e.g. adhesives.  Semi-permanent fasteners do permit removal, e.g. screws.

Sunday, 28 May 2006

Automatic Assembly (10/16)

The range of plastic snap-in fasteners is classed as permanent or semi-permanent, as they can be removed with the aid of special tools.  It is not feasible to repair products with these permanent joints.  This leads to 'throw away' products, upon a fault occurring, unless self-contained sub-assemblies are used in standard modules.

Parts integration dictates that groups of components should, where possible, be manufactured as a single part by chip-less forming, e.g. precision die-casting, precision plastic moulding, powder metallurgy, investment casting, fine blanking, and high energy rate forming. These methods produce, in one single operation, the features of a number of parts without the use of fasteners.  Also, features for identification by the automatic bowl feeder tooling can be cast into the part.

Each part in an assembly serves a purpose with the aid of functional features.  The designer should group a number of parts together to form a single part with multi-functional features.

Create a precedence diagram for the assembly operations. Identify redundant areas of the operation and incorporate the functional parts of these operations into other operations.

Saturday, 27 May 2006

Automatic Assembly (11/16)

Ensure that each part is fully and correctly specified for dimension, function, quality, material, and shape.  Don't accept supplier parts outside specification.  These parts may be incorrectly accepted on the grounds that they perform the same function, and yet they may not be acceptable to the automatic feeder tooling, workhead or fixturing.

Minimize the variation in component and product designs.  Minimizing the variation in part designs reduces the numbers of feeders used and enables standardisation of gripper and fixture design.  Minimizing variations reduces changeover times and enables standard magazines and packs to be used.

An existing product will be designed for production by manual assembly.  It is unlikely that an automatic assembly system will accept the part designs for a product currently assembled manually. The product needs to be re-designed to suit the machine principles, e.g. positional accuracy of a pick and place unit.

Friday, 26 May 2006

Automatic Assembly (12/16)

If a part-built assembly needs moving during assembly then problems arise if all parts are not located.  During manual assembly of a product, the operations are structured so that transportation only occurs with stable assemblies. This is achieved by assigning two or more parts to the assembly worker, or enough parts that are required, to create a stable structure. The operative, using two hands, holds the unstable part whilst assembling the part required to complete the operation.  An example of an operation such as this is where an assembly worker holds down a spring with one hand, prior to assembly of a spring retainer with the other hand.  This type of operation is difficult to perform automatically and should be re-designed so that each part is self-locating.

Design the product with many sub-assemblies.  Each sub-assembly should be common to all product styles.  Product variation can then be created in the final assembly of the product.  Sub-assembly work centres give a greater overall efficiency of the assembly system, in conjunction with buffer storage.  This is achieved by using a free transfer line or by intermediate storage systems.

The feeding of a part to an automatic workhead is by components in bulk random orientation or structured orientation.  Methods of feeding are usually determined by the part characteristics and required feed rate.  All feeders are classified as being; automatic, magazine, final parts forming stage, or manual.  It is uneconomic, or impossible, to feed certain parts automatically and these are not fed by automatic feeders. Flexible gaskets, open ended springs and acute angled cones are examples of such parts.  Large parts, parts having no symmetry, and delicate parts (e.g. with print face) cannot be fed automatically, but may be fed by magazines.  Relatively simple parts, with a degree of symmetry or definite asymmetry, can be fed by automatic feeders.

Thursday, 25 May 2006

Automatic Assembly (13/16)

The most common form of automatic parts feeder is the vibratory bowl feeder. Other automatic feeders include the hopper feeder, centrifugal hopper, barrel hopper, magnetic feeder, and elevating hopper feeder. These parts feeders are specialized and the feeding mechanism is designed to handle unique parts. The device which converts the bulk random orientation of parts into flow of orientated parts in these automatic feeders often takes the form of a fork or blade.

The movement of parts in a vibratory bowl feeder is created by a drive unit which induces vertical and angular vibration to the bowl.  Parts are momentarily caused to leave the track surface during the vertical phase of vibration.  The angular vibration moves the track and the part falls onto a track portion beyond the initial position of the part.  Incorrectly orientated parts are rejected at the bowl tooling stage by passive devices and returned to the bowl base. These mechanisms of operation make the bowl feeder unsuitable for automatic feeding of parts dosed with viscous fluids because the vertical vibrational force is not sufficient to lift the part from the track, due to the adhesion of the viscous fluid.  The continuous rejection of parts at the tooling stage and the vibration of the track against the part tends to damage delicate parts. Parts with print faces and delicate projections are particularly affected.

Wednesday, 24 May 2006

Automatic Assembly (14/16)

Once it has been determined that a part can be fed by a bowl feeder, the remaining consideration is the maximum feed rate that can be obtained from the feeder.  The feed rate of the part must be within the cycle time of the complete assembly operation.  This rate, for a given conveying velocity, depends upon the physical size of the component and the features for orientation.  For parts symmetrical about all axes, e.g. cubes, spheres, every part will leave the feeder 'first time through', or 100% of the parts will leave the bowl because no tooling is required.  This 'first time through' rate is the measure of a part's efficiency at being fed automatically.  These parts will always be correctly orientated and ready for insertion.  Components having little symmetry have a much lower tooling efficiency and it can be difficult to achieve the required feed rate using one bowl feeder with passive tooling.

Parts feeding is the most difficult area of assembly automation.  If the part can be automatically fed and orientated to the workhead then it can usually be assembled.

The final parts forming stage can be used as a method for feeding difficult parts.  The parts are usually fed in bulk on strips and pressed or guillotined, before being fed to the workheads.  The strips are easy to handle and orientation can be better controlled.

Tuesday, 23 May 2006

Automatic Assembly (15/16)

Parts are orientated by tooling, inside or outside of the bowl feeder.  In-bowl tooling tends to be passive and relies on the probability of a correctly orientated part moving along the conveyor track.  Incorrectly orientated parts are detected by the bowl tooling and deflected back to the bowl base.  Correctly orientated parts are accepted by the tooling and presented to the workhead.  Active tooling accepts parts in more than one orientation and re-orientates them correctly for the workhead.  This method of tooling gives a 100% 'first time through' rate and can be used outside the bowl.

It's important to make parts as symmetrical as possible.  Higher feed rates are obtained with parts having greater degrees of symmetry.  This is achieved by duplicating non-productive features.  If it is not feasible to make a part symmetrical then it must be designed to be definitely asymmetrical.  Features which are too small to be detected by bowl tooling must be exaggerated, for them to be detected.  The use of cylinders having a length to diameter ratio of unity and rectangles having slightly dis-similar sides should be re­designed to give greater asymmetry.

Monday, 22 May 2006

Automatic Assembly (16/16)

The most efficient insertion processes are those from vertically above, in a straight line movement.  Most assembly processes take this form.  If this is not the case with a part then it should be examined to see if the action can be simplified.  Simple insertion processes need low cost workheads.  This is because more complex operations require more degrees of freedom.  Each degree of freedom needs an individual pneumatic, hydraulic, or DC Servo motor and, therefore, the cost increases accordingly.  The cycle time of the operation also increases.


Increases in productivity can be realized by re-designing an existing product for automatic assembly.  Component re-design is more beneficial than assembly system re-design. All of the design considerations mentioned are related to the three main rules for design for automatic assembly :

1.Use the least number of parts
2.Obtain low feeding and orientation difficulty levels
3.Use simple insertion operations