The paper considers a relevant problem "Computer simulation of robotic device components in manufacturing on a 3D printer" and highlights the problem of computer simulation based on the cognitive programming technology of robotic device components. The paper subject is urgent because computer simulation of force-torque and accuracy characteristics of robot components in terms of their manufacturing properties and conditions from polymeric and metallic materials is of paramount importance for programming and manufacturing on the 3D printers. Two types of additive manufacturing technologies were used:

1. FDM (Fused deposition modeling) - layered growth of products from molten plastic strands;

2. SLM (Selective laser melting) - selective laser sintering of metal powders, which, in turn, create:

• conditions for reducing the use of expensive equipment;

• reducing weight and increasing strength through optimization of  the lattice structures when using a bionic design;

• a capability to implement mathematical modeling of individual components of robotic and other devices in terms of appropriate characteristics;

• a 3D printing capability to create unique items, which cannot be made by other known methods.

The paper aim was to confirm the possibility of ensuring the strength and accuracy characteristics of cases when printing from polymeric and metallic materials on a 3D printer. The investigation emphasis is on mathematical modeling based on the cognitive programming technology using the additive technologies in their studies since it is, generally, impossible to make the obtained optimized structures on the modern CNC machines.

The latter allows us to create a program code to be clear to other developers without cost, additional time for development, adaptation and implementation.

Year by year Russian companies increasingly use a 3D-print system in mechanical engineering, aerospace industry, and for scientific purposes. Machines for the additive production, well integrated in the production cycle, allow not only to reduce costs and save time, but also to begin performing more complex tasks.

The ability for quick understanding the program code written by other users and adaptation to other projects creates the preconditions for a good economic effect, strength through the use of already existing projects or models of standard forms. This is possible because, based on the data on three-dimensional detail layout scanning, the 3D technology allows us to have a digital copy to be adapted for further implementation on the 3D-printer in any potential project, taking into account the specification of each project.

Analysis of these problems gives impetus for further research in the field of mechanical engineering and robotics technology.

Conclusion: A process of simulation and bionic design of the case has been done; fixation was across frontal cross-section. Margin of safety for the case (option 1) made from metallic powder is sufficient. As seen from the two proposed options of the device case the first option is more preferable since there is a collapse potential in fixing points of the option 2 made from the ABC plastic.

The work deals with modeling methods of frame structures, used in rocket technology. To estimate their inertia, stiffness and strength properties it is necessary to create its calculation model. Currently, a finite element method is generally used. Various finite elements and combinations thereof may constitute such models. It is obvious that parameters of design under consideration, resulting from calculation, depend on the selected calculation model.

To simulate the properties of a longitudinal set of elements (stringers and longerons), as well as braces a girder (beam) idealization is offered, because their transverse dimensions are substantially smaller than the longitudinal ones. The elements of a transverse power set (frames) is a plated arch-shaped construction of variable sections with holes. Weld tubes of different diameters reinforce the holes in plates. There are cutouts in the vertical plates to mount the stringers. It is impossible to take completely into consideration the mentioned features in the frame idealization of bulkheads. Therefore, it seems more appropriate to use plate-frames for modeling the bulkheads.

. The paper discusses two possible models of the bulkhead to show their stiffness properties, namely frame and plate. Describes strengths and shortcomings of represented bulkhead design schemes. Based on the test calculations shows which errors may be possible in calculating the strength properties of the bulkhead under consideration when using a frame model. Gives recommendations on the proper use of potential bulkhead models to estimate a general and local strength of the bulkhead structure.

The SADAS software package developed at the Department SM8 in BMSTU was used to create the bulkhead models under study, perform their test calculations and analyze calculation results.

An important task of modern mathematical statistics with its methods based on the theory of probability is a scientific estimate of measurement results. There are certain costs under control, and under ineffective control when a customer has got defective products these costs are significantly higher because of parts recall.

When machining the parts, under the influence of errors a range scatter of part dimensions is offset towards the tolerance limit. To improve a processing accuracy and avoid defective products involves reducing components of error in machining, i.e. to improve the accuracy of machine and tool, tool life, rigidity of the system, accuracy of the adjustment. In a given time it is also necessary to adapt machine.

To improve an accuracy and a machining rate there, currently  become extensively popular various the in-process gaging devices and controlled machining that uses adaptive control systems for the process monitoring. Improving the accuracy in this case is compensation of a majority of technological errors. The in-cycle measuring sensors (sensors of active control) allow processing accuracy improvement by one or two quality and provide a capability for simultaneous operation of several machines.

Efficient use of in-cycle measuring sensors requires development of methods to control the accuracy through providing the appropriate adjustments. Methods based on the moving average, appear to be the most promising for accuracy control since they include data on the change in some last measured values of the parameter under control.