Design & Decision-Making for Mechanical Engineers
Engage in application of engineering knowledge—for it to translate into practical design decisions and develop engineering judgment to manage trade-offs while meeting design constraints.
What Design & Decision-Making Actually Means
Math problems may be David vs. Goliath, but the real battles of engineering are waged over imprecise information under tight deadlines with constrained budgets. Bolt vs weld? What to do with a "slobbery" dimension? What safety factor is just right? These issues don't have a unique answer based on mathematical calculations. Rather, they involve balanced decisions among cost, performance, manufacturability, reliability, and schedule.
Design and decision-making is where the fundamental knowledge is put against real world constraints. We can calculate the behavior of the system as the physics dictates, but then there are three good solutions. The challenge is deciding which solution best meets the company's needs, and defend that decision. This workshop bridges the gap between knowing the design principles and executing the design, whether that is product development, manufacturing, quality, or consulting.
Engineering Judgment: The Core Skill
Engineering judgment is making good technical decisions in the presence of uncertainty or incompleteness of information. Such judgments cannot always be based on complete data and necessarily require trade-offs, a sense of patterns, and good calibration. Engineering judgment is a skill that can be learned by studying successful examples and improving one's own judgments through feedback.
Judgment develops through three capabilities:
→ The trade-offs: Recognizing that improving on one aspect of a design will often degrade another. Cost vs. safety. Tight tolerances vs. manufacturability. Performance vs. reliability.
→ Managing uncertainty: Making progress despite missing or unknown information by using estimation and appropriate safety margins instead of waiting for more data that may never be available.
→ Learning from outcomes: Using outcomes to test mental models by learning from successes and failures, distinguishing between actual success or failure and mere luck, and using failures as opportunities to revise and improve mental models rather than defending bad decisions.
Thinking Like an Engineer section covers the basic thinking skills of an engineer such as problem definition, estimating, understanding limitations. Some examples of how these thinking skills are applied to engineering decision making for project development are included such as: process selection, tolerance levels, safety factors, self evaluation.
Where This Fits: Knowledge to Action
Fundamentals give you the knowledge base—what phenomena exist, how systems behave, why things fail. Specializations show how fundamentals apply in domains like FEA or robotics. Design & Decision-Making sits between—it's where general knowledge becomes specific action.
Cyclic loading will cause fatigue, but is that an issue for How Things Fail in your design? Tighter tolerances in your Engineering Drawings cost money, but are they worth it for specific dimensions? This section shows you how to make design decisions by connecting abstract engineering knowledge to your specific design. Not sure where to focus your studies? Our career assessment can help identify any knowledge gaps.
Six Core Topics
These are real issues that engineers face and make decisions about every day. Each of these addresses how to make design decisions when implementing system requirements.
Engineering Design Process
Design evolution. How do you iterate upon your design? Through what means do you prototype and test your work? When do you finally call a design done? Reality rarely follow the textbook flowchart. Design is about making deals with several constraints to achieve a greater good.
Load Cases & Assumptions
This is a series based on the ins and outs of building a solid design analysis, the first of which is defining what your design must survive. Loads and boundary conditions, assumptions that make or break your analysis.
Failure Modes & Design for Failure
Looking ahead to how things will break – anticipating failure points and designing things so that they fail safely or don't break in the first place.
Tolerances & Fits
Understanding why parts won't assemble can be critical to optimizing product development. Part interfaces require specific design tolerances that must be balanced across three critical factors: product cost, functional performance and manufacturability. When to tighten up the design and when to relax it.
Design for Manufacturing (DFM)
Design for manufacture, ensuring that what you design can actually be made, considering manufacturing constraints, appropriate manufacturing processes and avoiding costly redesign.
Reviewing Your Own Design
Catching mistakes before they cost too much: systematic self review techniques that find errors others don't, and boost your confidence.
Real-World Scenarios: When Design Decisions Matter
Every decision in engineering carries consequences: wasted budget, failed components, delayed projects, and crushed credibility. But junior engineers don't know the better path because no one ever taught them a simple, universal framework to guide their choices. Let's change that.
Scenario 1: Bolts vs. Welds—The Assembly Decision
The situation: The design challenge is to select a method to assemble a structural bracket intended for field assembly. Analysis indicates that both a weld and a bolt are acting structurally in the bracket design shown.
❌ The weak decision: Choosing welding because it is "stronger" – meanwhile the frame is powder coated and therefore welding it would be destructive, and the field installers are not even certified welders. Now you are in a grind to redesign under pressure. It may have added some sturdiness but it was a poor trade off in relation to all the other criteria that had to be considered.
✓ The strong decision: This decision is not just about the structural adequacy of the system. Factors such as manufacturing capabilities and field installation requirements must also be considered. The engineering plants where details are fabricated should be contacted early in the design process. Field assembly techniques require bolts because welding equipment is not provided on site. Designs that include bolts, specify the appropriate grades and torques, and justify their selection, have been developed and passed through the design review process smoothly.
Key lesson: Plan early and often your choices will limit your future options, incorporating stakeholders in the design process early on will prevent rework. See Engineering Design Process.
Scenario 2: Loose Tolerance Causing Vibration—The Specification Mistake
The situation: In designing a rotating assembly, its rotor is vibrating excessively during running tests. After detailed checks on alignment, balance and pre-load, critical measurement on the bearing bore diameter of the housing shows it to be in excess of tolerated limits and is allowing the bearing to move within the bore. The designer had selected a standard H7 tolerance based upon an example without appreciating its inappropriateness.
❌ The weak decision: The weak decision to copy tolerances and production methods from earlier designs without analyzing functional requirements for the part currently under development. The easy route of specifying loosen fits because "tight tolerances are expensive." Shipout with all the best intentions, and find the looseness to be a failure after the fact – only then frantically attempting to retrofits like using Loctite, poor quality press fits, design revisions.
✓ The strong decision: Understanding that bearings require precise fits to prevent unwanted micro-movements and ensuring correct rotating and non-rotating housing tolerances. Research the acceptable housing tolerance based on bearing loads and speeds. Decide on bearing fits from the start and document your decision making process. Remember tight tolerances cost money, but the cost of rework and delays is substantially higher.
Key lesson: Tolerances are design specifications - they are functional requirements, not an arbitrary number. Key dimensions need to be tightly controlled. See Tolerances & Fits.
Scenario 3: Overdesigned Part Increasing Cost—The Safety Factor Trap
The situation: I was designing a lever arm for a consumer product, which was specified to operate safely at 200 N load. The material had a yield stress of 250 MPa, and calculation showed that the worst case stress was 50 MPa, giving a margin of safety of 5. But my manager just blew me away when he asked why I designed it so "overstrength" that it was driving up cost and weight. I had always thought you couldn't be too safe, but this was a lesson that there is such a thing as too much safety.
❌ The weak decision: Arbitrarily use very high safety factors because "safer is better". Overdesign to an extent that makes products too expensive and too heavy. Competitors design more sound products, because they use safety factors of 1.5 to 2 for well-characterized, low-consequence consumer products, instead of space-age design margins which are applied to products that don't need them. Excessive conservatism is poor engineering, as opposed to cost effective engineering that makes it in the marketplace.
✓ Strong Decisions: Consensus States that the determination of the safety factors related to uncertainty involves an assessment of the consequences of failure and industry precedent. Deciding between a safety factor of 1.5 – 2, and 3-5 or more may involve a decision based on the ductility or brittleness of the chosen materials and the level of known or unknown loads. Document the basis of your decisions for these safety factors. This includes the level of certainty in the load data, the accuracy of the relevant material data, and the consequences of potential failure. Additionally, the relevant regulations must also be considered.
Key lesson: Safety factors are used to manage risk, and should be proportional to the true uncertainties and consequences for failure. See Failure Modes & Design for Failure.
Building Decision-Making Competence
I saw some very bad design decisions made by very knowledgeable engineers who didn't know how to apply what they knew. Making good decisions is not innate, it takes practice and it takes feedback. Study decision frameworks, study past decisions, get feedback from more senior engineers, and from manufacturing. With time, you build up a mental repository of experience that allows you to make quick, good decisions.
Don't wait for certainty: Make best bets with available data, justify your choices, and course correct based on consequences. Engineering that makes decisions well is more promostable than engineering that perfects a single feature. Freeze frame: Which is the engineer who stopped growing? The one who couldn't make a decision.
What Comes Next
However, the real meat is in the 6 individual topics which dive deeper into more specific design decisions such as the actual engineering design process, defining the proper load cases and assumptions for a product, thinking about potential failure modes and designing to avoid failure, and deciding on the proper tolerances and fits between components, in addition to designing for manufacturing and reviewing your own designs for errors.
Choose a topic which is currently relevant to your job. Each of the topics treated below are independent but relevant. Because real engineering decisions are interconnected, hopefully something here will ring true to you. If not, check out our section on Engineering Fundamentals to make sure you have a grasp of the underlying principles. Our career assessment quiz can also help you identify knowledge gaps and suggest specializations which will be a good fit for you.