Why can't a looped pipe network (with multiple paths between nodes) be solved by analyzing each pipe independently, the way a simple series or series-parallel pipe system can?
ABecause friction factors in pipes depend on temperature, which varies across a network
BBecause the flow in each pipe is unknown and depends on the flows in all other pipes through coupled continuity and energy constraints, making the equations interdependent
CBecause the Hardy-Cross method hasn't been applied yet, so no solution exists
DBecause real pipe networks always include pumps, which require separate analysis
In a simple series system, you know total flow and apply energy balance sequentially. In a parallel system, you know total flow splits and can use equal head-loss constraints to find the split. But in a looped network, flow can circulate around closed loops in either direction and at unknown magnitudes. The flow in pipe AB affects how much flow is available for pipes AC and BD, which in turn determines pressure drops that constrain other loops. Every pipe's flow depends on every other pipe's flow — the equations are fully coupled. There's no starting pipe where you can begin and work outward, so you must solve all equations simultaneously.
Question 2 Multiple Choice
During Hardy-Cross iteration, after applying flow corrections to all loops, you check and find that head-loss imbalances still exist in some loops. What is the correct interpretation and next step?
AThe network has no solution — real physical networks must always balance on the first try
BThe initial guesses violated continuity, which must be fixed before energy balance can be addressed
CThe solution hasn't converged yet — repeat the correction cycle, applying new ΔQ corrections based on the updated flows, until residuals become negligible
DThe friction factors (r values) are wrong and must be recalculated before proceeding
Hardy-Cross is an iterative method, not a direct solver. After each correction cycle, the flows are more accurate than before (they better satisfy energy balance) but not yet exact. The method converges because each correction reduces the residuals — it is a Newton-Raphson iteration on the energy balance equations, and Newton-Raphson typically reduces errors quadratically near the solution. Convergence is usually fast (3–5 iterations reduce residuals by orders of magnitude). The iteration terminates when all ΔQ corrections become smaller than the desired tolerance. Continuity is always maintained throughout, since the initial guess satisfies it and corrections are designed to preserve it.
Question 3 True / False
In Hardy-Cross iteration, the initial assumed flows must satisfy continuity (flow in equals flow out at every junction), even though they do not yet satisfy energy balance.
TTrue
FFalse
Answer: True
True, and this is a deliberate design feature of the method. Continuity (conservation of mass) is easy to satisfy by inspection — you can distribute arbitrary flows across the network as long as inflows equal outflows at every node. Energy balance (head loss summing to zero around loops) is what the iterations correct. The Hardy-Cross correction formula ΔQ = −ΔH/(n·Σ(r·|Q|ⁿ⁻¹)) is applied symmetrically to pipes shared between loops, ensuring that when you correct one loop, you don't violate continuity at any junction. Throughout the entire iteration, continuity is maintained exactly — only energy balance is progressively improved.
Question 4 True / False
Hardy-Cross is a specialized technique unique to pipe networks, fundamentally different from general numerical methods like Newton-Raphson, because pipe flow has special properties that require a custom algorithm.
TTrue
FFalse
Answer: False
False. Hardy-Cross is Newton-Raphson applied to the energy balance equations of a pipe network. The correction formula ΔQ = −ΔH/(n·Σ(r·|Q|ⁿ⁻¹)) is recognizable as −f(x)/f'(x): the numerator is the function (head-loss imbalance ΔH), and the denominator is its derivative with respect to flow (n·Σ(r·|Q|ⁿ⁻¹) = dΔH/dQ). What Hardy-Cross does is apply this Newton-Raphson update to each loop separately rather than globally — a simplification that works because loop corrections are approximately independent. Modern software uses the full global Newton-Raphson, which converges faster and handles pumps, valves, and pressure-controlled nodes, but it's the same underlying mathematics.
Question 5 Short Answer
What two physical conditions must be simultaneously satisfied in the final solution of a pipe network? Explain why satisfying only one of them is physically meaningless.
Think about your answer, then reveal below.
Model answer: The two conditions are: (1) continuity — at every junction, the sum of flow rates entering equals the sum leaving; and (2) energy conservation — around every closed loop, the net head loss sums to zero. Satisfying only continuity means flows are mass-balanced but violate thermodynamics: you'd be creating or destroying mechanical energy as fluid circulates around a loop, which is physically impossible in a real network. Satisfying only energy balance without continuity would require fluid to accumulate or disappear at junctions — also impossible. The two conditions together uniquely determine the flow distribution. Any physically realizable steady-state solution must satisfy both simultaneously, which is why iterative methods like Hardy-Cross must converge to a point where both constraints hold at once.
The key insight is that pipe network analysis is a constrained problem with two types of constraints from two different physical laws. Mass conservation gives one set of equations (continuity at nodes); energy conservation gives another set (loop energy balance). Only the intersection of solutions to both sets is physically valid. Hardy-Cross iterates from a mass-conservative starting point and progressively improves energy balance — maintaining one constraint throughout while converging on the other.